WiSe 24/25  
 Mathematics and...  
 Data Science  
 Course

Data Science

Data Science

0590b_MA120

  • Introduction to Profile Areas

    0590bA1.1

  • Statistics for Students of Data Science

    0590bA1.2
    • 19330401 Lecture
      Statistics for Data Science (Vesa Kaarnioja)
      Schedule: Mo 10:00-12:00, zusätzliche Termine siehe LV-Details (Class starts on: 2024-10-14)
      Location: A6/SR 032 Seminarraum (Arnimallee 6)

      Comments

      This course serves as an introduction to foundational aspects of modern statistical data analysis. Frequentist and Bayesian inference are presented from the perspective of probabilistic modelling.

      Details can be found on the website of the previous year.
    • 19330402 Practice seminar
      Practice Seminar Statistics for Data Sci (Vesa Kaarnioja)
      Schedule: Di 10:00-12:00 (Class starts on: 2024-10-15)
      Location: A7/SR 031 (Arnimallee 7)

  • Machine Learning for Data Science

    0590bA1.3
    • 19330101 Lecture
      Machine Learning for Data Science (Grégoire Montavon)
      Schedule: Di 16:00-18:00, Do 16:00-18:00, zusätzliche Termine siehe LV-Details (Class starts on: 2024-10-15)
      Location: T9/Gr. Hörsaal (Takustr. 9)

      Comments

      Qualifikationsziele:

      The course provides an overview of machine learning methods and algorithms for different learning tasks, namely supervised, unsupervised and reinforcement learning.

      In the first part of the course, for each task the main algorithms and techniques will be covered including experimentation and evaluation aspects.

      In the second part of the course, we will focus on specific learning challenges including high-dimensionality, non-stationarity, label-scarcity and class-imbalance.

      By the end of the course, you will have learned how to build machine learning models for different problems, how to properly evaluate their performance and how to tackle specific learning challenges.

      Inhalte

       Es werden Themen aus folgenden Gebieten behandelt:

       

      • Experiment Design
      • Sampling Techniques
      • Data cleansing
      • Storage of large data sets
      • Data visualization and graphs
      • Probabilistic data analysis
      • Prediction methods
      • Knowledge discovery
      • Neural networks
      • Support vector machines
      • Reinforcement learning and agent models
    • 19330102 Practice seminar
      Practice Seminar Machine Learning DatSci (Grégoire Montavon)
      Schedule: Mi 16:00-18:00 (Class starts on: 2024-10-16)
      Location: T9/SR 006 Seminarraum (Takustr. 9)

  • Programming for Data Science

    0590bA1.4
    • 19330313 Lab Seminar
      Programming for Data Science (Sandro Andreotti)
      Schedule: Mo 12:00-16:00 (Class starts on: 2024-10-28)
      Location: A3/SR 120 (Arnimallee 3-5)

      Comments

      Qualifikationsziele

      Die Studentinnen und Studenten haben ein tieferes Verständnis für Konzepte in der Programmierung mit einer höheren Programmiersprache (z. B. C/C++, Java oder Python).

      Inhalte:

      Einführung in verschiedene Arten von Programmiertechniken.

  • Advanced Topics in Data Management

    0089cA1.29
    • 19304801 Lecture
      Geospatial Databases (Agnès Voisard)
      Schedule: Di 14:00-16:00 (Class starts on: 2024-10-15)
      Location: T9/055 Seminarraum (Takustr. 9)

      Additional information / Pre-requisites

      Zielgruppe:

      Studierende im Masterstudiengang Voraussetzungen: Vorlesung: Einf. in Datenbanksysteme

      Comments

      The goal of this course is to acquire the background of spatial databases, the kernel of Geographic Systems. The major aspects that will be handled are: modeling and querying geospatial information, spatial access methods (SAMs), data representation, basic operations (mostly from computational geometry), and optimization. Insights into current applications such as location-based services (e.g., navigation systems) will also be given. Knowledge in databases is necessary. This course encompasses: formal lectures, exercises, as well as a practical project with PostGIS.
       

      Suggested reading

      Handouts are enough to understand the course.

      The following book will be mostly used: P. Rigaux, M. Scholl, A. Voisard.Spatial Databases - With Application to GIS. Morgan Kaufmann, May 2001. 432 p. (copies in the main library)
    • 19304802 Practice seminar
      Practice seminar for Geospatial Databases (Agnès Voisard)
      Schedule: Do 14:00-16:00 (Class starts on: 2024-10-17)
      Location: T9/Gr. Hörsaal (Takustr. 9)

  • Advanced Algorithms

    0089cA2.1
    • 19303501 Lecture
      Advanced Algorithms (László Kozma)
      Schedule: Di 10:00-12:00, Fr 10:00-12:00 (Class starts on: 2024-10-15)
      Location: T9/SR 006 Seminarraum (Takustr. 9)

      Additional information / Pre-requisites

      Target audience

      All Master and Bachelor students who are interested in algorithms.

      Prerequisites

      Basic familiarity with the design and analysis of algorithms.

      Comments

      This course will focus on the design and analysis of algorithms, with topics including:

      • general principles of algorithm design,
      • randomized algorithms,
      • dynamic programming,
      • flow problems on graphs,
      • amortized analysis and advanced data structures,
      • theory of NP-completeness,
      • approximation methods for hard problems,
      • other topics.

      Prerequisites are basic knowledge of algorithms and relevant mathematics. All Bachelor and Master students interested in advanced algorithmic techniques are welcome. Lectures are in English.

      Suggested reading

      • Cormen, Leiserson, Rivest, Stein: Introduction to Algorithms, 4th Ed. MIT Press 2022
      • Kleinberg, Tardos: Algorithm Design, Addison-Wesley 2005.
      • Sedgewick, Wayne: Algorithms, 4th Ed., Addison-Wesley 2016
    • 19303502 Practice seminar
      Practice seminar for Advanced Algorithms (László Kozma)
      Schedule: Fr 08:00-10:00, Fr 14:00-16:00 (Class starts on: 2024-10-18)
      Location: T9/046 Seminarraum (Takustr. 9)

  • Telematics

    0089cA3.5
    • 19305101 Lecture
      Telematics (Jochen Schiller)
      Schedule: Mo 12:00-14:00, Mi 10:00-12:00 (Class starts on: 2024-10-14)
      Location: T9/046 Seminarraum (Takustr. 9)

      Additional information / Pre-requisites

      Requirements: Basic understanding of computer networks, e.g., TI-III

       

      Comments

      This course addresses communication asp. The lecture addresses topics such as:

      • Basic background: protocls, services, models, communication standards;
      • Principles of communication engineering: signals, coding, modulation, media;
      • Data link layer: media access etc.;
      • Local networks: IEEE-Standards, Ethernet, bridges;
      • Network layer: routing and forwarding, Internet protocols (IPv4, IPv6);
      • Transport layer: quality of service, flow control, congestion control, TCP;
      • Internet: TCP/IP protocol suite;
      • Applications: WWW, security, network management;
      • New network concepts.

      In the supplementary exercise course the students will practically apply their knowledge.

      Suggested reading

      • Larry Peterson, Bruce S. Davie: Computernetze - Ein modernes Lehrbuch, dpunkt Verlag, Heidelberg, 2000
      • Krüger, G., Reschke, D.: Lehr- und Übungsbuch Telematik, Fachbuchverlag Leipzig, 2000
      • Kurose, J. F., Ross, K. W.: Computer Networking: A Top-Down Approach Featuring the Internet, Addi-son-Wesley Publishing Company, Wokingham, England, 2001
      • Siegmund, G.: Technik der Netze, 4. Auflage, Hüthig Verlag, Heidelberg, 1999
      • Halsall, F.: Data Communi-cations, Computer Networks and Open Systems 4. Auflage, Addison-Wesley Publishing Company, Wokingham, England, 1996
      • Tanenbaum, A. S.: Computer Networks, 3. Auflage, Prentice Hall, Inc., New Jersey, 1996
    • 19305102 Practice seminar
      Practice seminar for Telematics (Marius Max Wawerek)
      Schedule: Mo 16:00-18:00 (Class starts on: 2024-10-14)
      Location: T9/SR 006 Seminarraum (Takustr. 9)

  • Special Aspects of Data Science in Life Sciences

    0590bB1.4
    • 19328301 Lecture Cancelled
      Data Visualization (Claudia Müller-Birn)
      Schedule: -
      Location: keine Angabe

      Additional information / Pre-requisites

      https://www.mi.fu-berlin.de/en/inf/groups/hcc/teaching/winter_term_2021_22/course_data_visualization.html

      Comments

      The current rapid technological development requires the processing of large amounts of data of various kinds to make them usable by humans. This challenge affects many areas of life today, such as research, business, and politics. In these contexts, decision-makers use data visualizations to explain information and its relationships through graphical representations of data. This course aims to familiarize students with the principles, techniques, and methods in data visualization and provide practical skills for designing and implementing data visualizations.

      This course gives students a solid introduction to the fundamentals of data visualization with current insights from research and practice. By the end of the course, students will

      1. Be able to select and apply methods for designing visualizations based on a problem,
      2. know essential theoretical basics of visualization for graphical perception and cognition,
      3. know and be able to select visualization approaches and their advantages and disadvantages,
      4. be able to evaluate visualization solutions critically, and
      5. have acquired practical skills for implementing visualizations.

      This course is intended for students interested in using data visualization in their work and students who want to develop visualization software. Basic knowledge of programming (HTML, CSS, Javascript, Python) and data analysis (e.g., R) is helpful.

      In addition to participating in class discussions, students will complete several programming and data analysis assignments. In a mini-project, students work on a given problem. Finally, we expect students to document and present their assignments and mini-project in a reproducible manner.

      Please note that the course will focus on how data is visually coded and presented for analysis after the data structure and its content are known. We do not cover exploratory analysis methods for discovering insights in data are not the focus of the course.

      Suggested reading

      Textbuch

      Munzner, Tamara. Visualization analysis and design. AK Peters/CRC Press, 2014.

       

      Zusätzliche Literatur

      Kirk, Andy: Data visualisation: A handbook for data driven design. Sage. 2016.

      Yau, Nathan: Visualize This: The FlowingData Guide to Design, Visualization, and Statistics. Wiley Publishing, Inc. 2011.

      Spence, Robert: Information Visualization: Design for Interaction. Pearson. 2007.
    • 19333001 Lecture
      Cybersecurity and AI I: Privacy, Biometry, Certification (Gerhard Wunder)
      Schedule: Di 12:00-14:00 (Class starts on: 2024-10-15)
      Location: T9/SR 006 Seminarraum (Takustr. 9)
    • 19336801 Lecture
      Integrative analysis and including prior knowledge for data in the life sciences (Katharina Baum, Pauline Hiort, Pascal Iversen)
      Schedule: Mi 10:00-12:00 (Class starts on: 2024-10-16)
      Location: A6/SR 007/008 Seminarraum (Arnimallee 6)

      Comments

      Especially in the life sciences, data of different origins are often available for a question, and researchers already have prior knowledge, for example on dynamic aspects, or on spatial or regulatory relationships between entities. This course deals with analysis methods that can combine different data and prior knowledge. For example, we discuss how to link continuous and categorical data in mixed models, but also network integration, or multi-factorial matrix multiplication. A focus topic will deal with various approaches to informed machine learning such as graph-neural networks, transfer learning or current research methods such as simulation-based pre-training. The focus here is explicitly not on the processing of images, but on tabular or other data types. This course will be offered in English.
    • 60102001 Lecture
      Methods for clinical trials (Pimrapat Gebert, Maja Krajewska)
      Schedule: Fr 14:00-16:00 (Class starts on: 2024-10-18)
      Location: A6/SR 032 Seminarraum (Arnimallee 6)

      Comments

      In this course, we introduce and discuss statistical methods and study design applied in clinical trials. Basic design aspects like randomization, blinding, the definition of control groups and endpoints will be discussed as well as different study types such as efficacy trials, equivalence and bioequivalence trials, phase I, II and III trials and principles of meta analyses. The related statistical models and test will be introduced as well. The aim of the lectures is to learn about biometrical thinking in the context of clinical trials which includes the application of statistical methods, but also critical thinking on the experimental setting.
    • 60102301 Lecture
      Statistical Methods for Small Sample Sizes (Frank Konietschke)
      Schedule: Mi 14:00-16:00 (Class starts on: 2024-10-16)
      Location: A6/SR 031 Seminarraum (Arnimallee 6)

      Comments

      In this course, we introduce and discuss statistical inference methods for analyzing trials with small sample sizes. We hereby explore the impact of the standard assumption “N is large” and try to find an answer to the question “what means large?” The inference methods will cover estimation of treatment effects, confidence interval computations and hypothesis testing in both parametric and nonparametric models. Rank tests, bootstrap and permutation methods will be investigated in detail as approximation methods. This class aspires to learn about modern statistical tools that were designed to make accurate conclusions when sample sizes are rather small.

       
    • 19328302 Practice seminar Cancelled
      Data Visualization (Claudia Müller-Birn)
      Schedule: -
      Location: keine Angabe
    • 19333002 Practice seminar
      Practice seminar for Cybersecurity and AI I (Gerhard Wunder)
      Schedule: Mo 14:00-16:00 (Class starts on: 2024-10-14)
      Location: T9/SR 006 Seminarraum (Takustr. 9)
    • 19336802 Practice seminar
      Integrative analysis of biomedical data tutorials (Katharina Baum)
      Schedule: Fr 10:00-12:00 (Class starts on: 2024-10-18)
      Location: T9/046 Seminarraum (Takustr. 9)
    • 60102002 Practice seminar
      Practice seminar for Methods for clinical trials (Pimrapat Gebert, Maja Krajewska)
      Schedule: Fr 16:00-18:00 (Class starts on: 2024-10-18)
      Location: A6/SR 032 Seminarraum (Arnimallee 6)
    • 60102302 Practice seminar
      Practice seminar for Statistical Methods for Small Sample Sizes (Frank Konietschke)
      Schedule: Mi 16:00-18:00 (Class starts on: 2024-10-16)
      Location: A6/SR 031 Seminarraum (Arnimallee 6)

  • Current Research Topics: Data Science in the Life Sciences

    0590bB1.5
    • 19328301 Lecture Cancelled
      Data Visualization (Claudia Müller-Birn)
      Schedule: -
      Location: keine Angabe

      Additional information / Pre-requisites

      https://www.mi.fu-berlin.de/en/inf/groups/hcc/teaching/winter_term_2021_22/course_data_visualization.html

      Comments

      The current rapid technological development requires the processing of large amounts of data of various kinds to make them usable by humans. This challenge affects many areas of life today, such as research, business, and politics. In these contexts, decision-makers use data visualizations to explain information and its relationships through graphical representations of data. This course aims to familiarize students with the principles, techniques, and methods in data visualization and provide practical skills for designing and implementing data visualizations.

      This course gives students a solid introduction to the fundamentals of data visualization with current insights from research and practice. By the end of the course, students will

      1. Be able to select and apply methods for designing visualizations based on a problem,
      2. know essential theoretical basics of visualization for graphical perception and cognition,
      3. know and be able to select visualization approaches and their advantages and disadvantages,
      4. be able to evaluate visualization solutions critically, and
      5. have acquired practical skills for implementing visualizations.

      This course is intended for students interested in using data visualization in their work and students who want to develop visualization software. Basic knowledge of programming (HTML, CSS, Javascript, Python) and data analysis (e.g., R) is helpful.

      In addition to participating in class discussions, students will complete several programming and data analysis assignments. In a mini-project, students work on a given problem. Finally, we expect students to document and present their assignments and mini-project in a reproducible manner.

      Please note that the course will focus on how data is visually coded and presented for analysis after the data structure and its content are known. We do not cover exploratory analysis methods for discovering insights in data are not the focus of the course.

      Suggested reading

      Textbuch

      Munzner, Tamara. Visualization analysis and design. AK Peters/CRC Press, 2014.

       

      Zusätzliche Literatur

      Kirk, Andy: Data visualisation: A handbook for data driven design. Sage. 2016.

      Yau, Nathan: Visualize This: The FlowingData Guide to Design, Visualization, and Statistics. Wiley Publishing, Inc. 2011.

      Spence, Robert: Information Visualization: Design for Interaction. Pearson. 2007.
    • 19333001 Lecture
      Cybersecurity and AI I: Privacy, Biometry, Certification (Gerhard Wunder)
      Schedule: Di 12:00-14:00 (Class starts on: 2024-10-15)
      Location: T9/SR 006 Seminarraum (Takustr. 9)
    • 19336801 Lecture
      Integrative analysis and including prior knowledge for data in the life sciences (Katharina Baum, Pauline Hiort, Pascal Iversen)
      Schedule: Mi 10:00-12:00 (Class starts on: 2024-10-16)
      Location: A6/SR 007/008 Seminarraum (Arnimallee 6)

      Comments

      Especially in the life sciences, data of different origins are often available for a question, and researchers already have prior knowledge, for example on dynamic aspects, or on spatial or regulatory relationships between entities. This course deals with analysis methods that can combine different data and prior knowledge. For example, we discuss how to link continuous and categorical data in mixed models, but also network integration, or multi-factorial matrix multiplication. A focus topic will deal with various approaches to informed machine learning such as graph-neural networks, transfer learning or current research methods such as simulation-based pre-training. The focus here is explicitly not on the processing of images, but on tabular or other data types. This course will be offered in English.
    • 60102301 Lecture
      Statistical Methods for Small Sample Sizes (Frank Konietschke)
      Schedule: Mi 14:00-16:00 (Class starts on: 2024-10-16)
      Location: A6/SR 031 Seminarraum (Arnimallee 6)

      Comments

      In this course, we introduce and discuss statistical inference methods for analyzing trials with small sample sizes. We hereby explore the impact of the standard assumption “N is large” and try to find an answer to the question “what means large?” The inference methods will cover estimation of treatment effects, confidence interval computations and hypothesis testing in both parametric and nonparametric models. Rank tests, bootstrap and permutation methods will be investigated in detail as approximation methods. This class aspires to learn about modern statistical tools that were designed to make accurate conclusions when sample sizes are rather small.

       
    • 19328302 Practice seminar Cancelled
      Data Visualization (Claudia Müller-Birn)
      Schedule: -
      Location: keine Angabe
    • 19333002 Practice seminar
      Practice seminar for Cybersecurity and AI I (Gerhard Wunder)
      Schedule: Mo 14:00-16:00 (Class starts on: 2024-10-14)
      Location: T9/SR 006 Seminarraum (Takustr. 9)
    • 19336802 Practice seminar
      Integrative analysis of biomedical data tutorials (Katharina Baum)
      Schedule: Fr 10:00-12:00 (Class starts on: 2024-10-18)
      Location: T9/046 Seminarraum (Takustr. 9)
    • 60102302 Practice seminar
      Practice seminar for Statistical Methods for Small Sample Sizes (Frank Konietschke)
      Schedule: Mi 16:00-18:00 (Class starts on: 2024-10-16)
      Location: A6/SR 031 Seminarraum (Arnimallee 6)

  • Seminar: Data Science in the Life Sciences (for Master's Students)

    0590bB1.6
    • 19333611 Seminar
      Seminar Deep Learning for biomedical applications (Vitaly Belik)
      Schedule: Mo 16:00-18:00 (Class starts on: 2024-10-14)
      Location: T9/051 Seminarraum (Takustr. 9)

      Comments

      Recent developments in the area of machine learning due to availability of data and computational power promise to revolutionize almost every area of science. The driving technology behind this advancement is deep learning – a machine learning technology based on artificial neural networks consisting of many layers. Deep learning is capable of processing huge amount of data of different nature and already outperforming humans in many decision-making tasks. Biomedical research became now a source of large heterogeneous data, i.e. images, video, activity sensors, omics and text data. Leveraging the opportunities of this deep learning technology in the biomedical field requires particular set of skills combining thorough knowledge of necessary algorithms, specifics of biomedical data and designated programming tools. In this course we aim to offer students with background in computer science an opportunity to acquire the above skills to be able to deploy deep learning technology with a focus on biomedical applications. The course is structured as a seminar, where students under extensive guidance of instructors read fundamental books and recent research articles on deep learning, learn necessary programming tools, and produce their own implementations of computational pipelines in case studies using already published or original data. Starting from fundamental aspects of deep learning we aim to cover its applications to e.g. image data, time series data, text data, complex networks.

      Suggested reading

      [1] Andresen N, Wöllhaf M, Hohlbaum K, Lewejohann L, Hellwich O, Thöne- Reineke C, Belik V, Towards a fully automated surveillance of well-being status in laboratory mice using deep learning: Starting with facial expres- sion analysis. Plos One, 15(4):e0228059, (2020) https://doi.org/10.1371/ journal.pone.0228059

      [2] Jarynowski A, Semenov A, Kamiński M, Belik V. Mild Adverse Events of Sputnik V Vaccine in Russia: Social Media Content Analysis of Telegram via Deep Learning. J Med Internet Res 2021;23(11):e30529 https://doi.org//10.2196/30529
    • 19334617 Seminar / Undergraduate Course
      Seminar/Proseminar: Multi-Agent Reinforcement Learning (Tim Landgraf)
      Schedule: Mo 10:00-12:00 (Class starts on: 2024-10-14)
      Location: A6/SR 007/008 Seminarraum (Arnimallee 6)

      Comments

      This seminar provides an exploration of large language models (LLMs), covering both foundational concepts and the latest advancements in the field. Participants will gain a comprehensive understanding of the architecture, training, and applications of LLMs, based on seminal research papers. The course will be organised as a journal club: students present individual papers, which are then discussed in the group to make sure we all get the ideas presented.

      ### Potential Topics

         - Neural networks and deep learning basics

         - Sequence modeling and RNNs (Recurrent Neural Networks)

         - Vaswani et al.'s "Attention is All You Need" paper

         - Self-attention mechanism

         - Multi-head attention and positional encoding

         - GPT-1: Radford et al.'s pioneering work

         - GPT-2: Scaling and implications

         - GPT-3: Architectural advancements and few-shot learning

         - BERT (Bidirectional Encoder Representations from Transformers)

         - T5 (Text-To-Text Transfer Transformer)

         - DistilBERT and efficiency improvements

         - Mamba:l and other SSMs: Design principles and performance

         - Flash Attention et al: Improving efficiency and scalability

         - Training regimes and resource requirements

         - Fine-tuning and transfer learning

      - Emergence of new capabilities
    • 19335011 Seminar
      Seminar: Networks, dynamic models and ML for data integration in the life sciences (Katharina Baum, Pauline Hiort, Pascal Iversen)
      Schedule: Fr 12:00-13:30 (Class starts on: 2024-07-26)
      Location: T9/137 Konferenzraum (Takustr. 9)

      Comments

      Research seminar of the group Data Integration in the Life Sciences (DILiS). Also open for seminar participation in the Master's program, online participation possible. Please refer to the current schedule on the whiteboard!

      The seminar offers space for the discussion of advanced and integrative data analysis techniques, in particular presentations and discussion of ongoing or planned research projects, news from conferences, review and discussion of current literature and discussion of possible future teaching formats and content, and presentations, as well as final presentations on theses or project seminars. The seminar language is mostly English. Interested students are welcome to attend and drop in without obligation or present a topic of their own choice of interest to the working group as in a usual seminar. Please note: Individual dates may be canceled or postponed. Please contact me in case of questions (katharina.baum@fu-berlin.de)!
    • 19336311 Seminar
      Visualization for Artificial Intelligence Explainability (Georges Hattab)
      Schedule: Mi 10:00-12:00 (Class starts on: 2024-10-23)
      Location: A3/019 Seminarraum (Arnimallee 3-5)

      Comments

      As AI systems grow more powerful, there is an increasing need to make these complex "black box" models interpretable and explainable. This seminar explores how data visualization techniques can provide crucial insights into how AI models operate and arrive at their outputs. Cutting-edge methods like saliency maps, decision trees, and dimensionality reduction visualizations allow us to peer inside deep neural networks and understand what factors they are considering.

       

      The seminar also covers visualization literacy - effectively communicating AI explainability visualizations to different stakeholders. Case studies highlight best practices for visualizing model behavior, evaluating fairness, and instilling appropriate levels of trust. Attendees will gain an understanding of how visualization can demystify AI, foster transparency, and enable real-world deployment of these systems in high-stakes domains.
    • 19336717 Seminar / Undergraduate Course
      Graph-neural networks in the life sciences and beyond (Katharina Baum, Pauline Hiort, Pascal Iversen)
      Schedule: Di 12:00-14:00 (Class starts on: 2024-10-15)
      Location: A6/SR 009 Seminarraum (Arnimallee 6)

      Comments

      Complex data can often be naturally modeled as a graph. Graphs or networks describe the interaction between objects and are an effective tool to represent systems in many applications. Graph neural networks are neural networks that directly input graphs and have recently emerged as a powerful tool to analyze networks and to predict properties of nodes and connections.

      This seminar offers an in-depth exploration of Graph Neural Networks (GNNs) and their applications across various domains, with a particular emphasis on the life sciences and biomedicine. We will begin by discussing the fundamental concepts and architectures of GNNs, including graph convolutional networks (GCNs) and graph attention networks (GATs). Applications that are discussed include protein-protein interaction networks, drug discovery and personalized medicine. Students will read and present research papers and participate in critical discussions.

      The language of this seminar is planned to be English. The students are encouraged to present and discuss in English, but contributions in German are also possible.
    • 19337211 Seminar
      Representation Learning (Georges Hattab)
      Schedule: Mi 14:00-16:00 (Class starts on: 2024-10-23)
      Location: A3/019 Seminarraum (Arnimallee 3-5)

      Comments

      While traditional feature engineering has been successful, modern machine learning increasingly relies on representation learning - automatically discovering informative features or representations from raw data. This seminar dives into advanced neural network-based approaches that learn dense vector representations capturing the underlying explanatory factors in complex, high-dimensional datasets.

       

      The seminar will cover techniques like autoencoders, variational autoencoders, and self-supervised contrastive learning methods that leverage unlabeled data to learn rich representations. You'll learn about properties of effective learned representations like preserving locality, handling sparse inputs, and disentangling underlying factors. Case studies demonstrate how representation learning enables breakthrough performance on tasks like image recognition and natural language understanding. You'll gain insights into interpreting these learned representations as well as their potential and limitations.
    • 19402911 Seminar
      Journal Club Computational Biology (Knut Reinert)
      Schedule: Mo 14:00-16:00 (Class starts on: 2024-10-14)
      Location: T9/053 Seminarraum (Takustr. 9)

      Additional information / Pre-requisites

      Open for:

      Master and PhD students

      Comments

      Content:

      In this seminar we will present original work in Computational biology as well as progress reports from PhD students. Master students will either be assigend a paper, or present their MSc thesis plans and results, or report about their research internship. Credits are only awarded for the presentation of papers.

      Please sign up on the Whiteboard (open "Site Browser" and look forJournal Club).

      Suggested reading

      aktuelle Publikationen aus der Forschung
    • 19404611 Seminar
      Open science, data handling and ethical aspects in bioinformatics (Thilo Muth)
      Schedule: Do 14:00-16:00 (Class starts on: 2024-10-17)
      Location: T9/051 Seminarraum (Takustr. 9)

      Comments

      The main objectives of this seminar are (1) to introduce fundamental methods of data handling in bioinformatics research with a particular focus on lab management systems, containerized frameworks and software workflow systems, (2) to provide a general overview of open science, sustainable software development, information security and data protection and (3) to discuss ethical issues, chances and risks with regard to fundamental research and personalized medicine.

      Planned topics are as follows:

      • Open vs. closed Science: chances and risks of open science. Open access and traditional publishing.
      • Sustainable software development and reproducible research: open data and open source, code repositories, maintainability of software and code in the sciences
      • Laboratory information management systems: sample management, integration of instruments and application, data exchange
      • Bioinformatic workflows: Galaxy, Snakemake and KNIME
      • Containerized frameworks: Advantages of using Docker, Bioconda etc.
      • Information security and data protection: securing personalized data, relevance for data analyses in research, efficient handling guidelines of data security, potential issues with open science
      • Ethical issues, chances and risks of omics research: immense opportunities when manipulating genome information, novel ethical questions need to be asked, „right not to know“, danger of discrimination against individuals or groups based on genetic differences

      Organisational note: During the first course at the beginning of the semester, the aforementioned topics will be briefly introduced and course material (e.g. relevant publications) will be provided. In the second course, topics will be assigned to the attendees of the seminar.
    • 19405911 Seminar
      Biochemical networks and disease (Jana Wolf)
      Schedule: Mi 12:00-14:00 (Class starts on: 2024-10-16)
      Location: T9/051 Seminarraum (Takustr. 9)

      Comments

      Molecular metabolic, signaling and gene-regulatory networks form complex networks that underly the normal physiological functioning of the cell. Various perturbations within these networks have been described in diseases. We will here use original papers to study and discuss how perturbations can be implemented in models and how they change the network characteristics. We will focus on dynamic models described by ordinary differential equations.
    • 19406411 Seminar
      Journal Club: Public Health Data Science (Max von Kleist)
      Schedule: Erster Termin: Mittwoch 23.Oct. Zeit: Mittwoch 10:00 (s.t.)-11:30 (Class starts on: 2024-10-23)
      Location: Online (https://rki.webex.com/rki/j.php?MTID=m032171b9d9eab45185ea259824052554 )

      Comments

      In this seminar, current research in the field of data-driven public health science, as well as the progress reports of PhD students and post-docs, will be presented. Master's students will present either an assigned journal article or their master's thesis, or they will report on their research internship. Credits will be awarded for article presentations only.

      Schedule: online, by arrangement

  • Selected Topics in Data Science in Life Sciences

    0590bB1.7
    • 19315401 Lecture
      Graph and Network Algorithms (Günther Rothe)
      Schedule: Mo 14:00-16:00, Fr 12:00-14:00 (Class starts on: 2024-10-14)
      Location: T9/SR 005 Übungsraum (Takustr. 9)

      Additional information / Pre-requisites

      Target Audience

      Masters students in Computer Science or Mathematics, advanced Bachelor students.

      Prerequisites

      "Advanced Algorithms" or a similar class

      Comments

      Graphs and networks are an important modeling tool for all kinds of relations in Computer Science and beyond, for example social networks, traffic networks, and so on. We will treat algorithmic problems that arise in this context:

      • analysis of networks
      • optimization in graphs
      • graph drawing

      Suggested reading

      Wird noch bekannt gegeben.
    • 19315402 Practice seminar
      Practice seminar for Graph and Network Algorithms (Mahmoud Elashmawi, Günther Rothe)
      Schedule: Mi 12:00-14:00 (Class starts on: 2024-10-16)
      Location: T9/053 Seminarraum (Takustr. 9)

  • Special Aspects of Data Science Technologies

    0590bB2.3
    • 19327201 Lecture
      Data compression (Heiko Schwarz)
      Schedule: Mo 14:00-16:00 (Class starts on: 2024-10-14)
      Location: T9/049 Seminarraum (Takustr. 9)

      Comments

      Data compression is a technology, which only enables a variety of applications in our information age. Even though the underlying technology is often hidden from the end user, we use data compression every day when we hear music, watch images and videos, or use applications on our smartphone.

      In this course, the fundamental and most often used approaches for data compression are introduced.  We discuss theoretical foundations as well as methods used in practice.

      The first part of the course deals with lossless compression, in which the original data can be reconstructed exactly. This part includes the following topics:

      • Unique decodability and prefix codes
      • Entropy and entropy rate as theoretical limits of lossless compression
      • Optimal codes, Huffman codes
      • Arithmetic coding
      • Lempel-Ziv coding
      • Linear prediction
      • Examples from text, image and audio compression

      In the second part of the course, we consider lossy compression, by which only an approximation of the original data can be reconstructed. This type of compression enables much higher compression rates and is the dominant form of compression for audio, image and video data. The second part of the course includes the following topics:

      • Scalar quantization, optimal scalar quantization
      • Theoretical limits of lossy compression: Rate distortion functions
      • Vector quantization
      • Predictive quantization
      • Transform coding
      • Examples from audio, image, and video compression

      Suggested reading

      • Sayood, K. (2018), “Introduction to Data Compression,” Morgan Kaufmann, Cambridge, MA.
      • Cover, T. M. and Thomas, J. A. (2006), “Elements of Information Theory,” John Wiley & Sons, New York.
      • Gersho, A. and Gray, R. M. (1992), “Vector Quantization and Signal Compression,” Kluwer Academic Publishers, Boston, Dordrecht, London.
      • Jayant, N. S. and Noll, P. (1994), “Digital Coding of Waveforms,” Prentice-Hall, Englewood Cliffs, NJ, USA.
      • Wiegand, T. and Schwarz, H. (2010), “Source Coding: Part I of Fundamentals of Source and Video Coding,” Foundations and Trends in Signal Processing, vol. 4, no. 1-2.
    • 19328601 Lecture
      Cryptocurrencies and Blockchain (Katinka Wolter, Justus Purat)
      Schedule: Di 10:00-12:00 (Class starts on: 2024-10-15)
      Location: T9/049 Seminarraum (Takustr. 9)

      Comments

      We will study the history, technology and applications of cryptocurrencies and blockchain.

      Suggested reading

      Bitcoin and Cryptocurrency Technologies: A Comprehensive Introduction, by Arvind Narayanan, Joseph Bonneau, Edward Felten, Andrew Miller, Steven Goldfeder
    • 19333001 Lecture
      Cybersecurity and AI I: Privacy, Biometry, Certification (Gerhard Wunder)
      Schedule: Di 12:00-14:00 (Class starts on: 2024-10-15)
      Location: T9/SR 006 Seminarraum (Takustr. 9)
    • 19327202 Practice seminar
      Practice seminar for Data Compression (Heiko Schwarz)
      Schedule: Mo 12:00-14:00 (Class starts on: 2024-10-14)
      Location: T9/049 Seminarraum (Takustr. 9)
    • 19328602 Practice seminar
      Practice Session on Cryptocurrencies (Justus Purat)
      Schedule: Do 10:00-12:00, zusätzliche Termine siehe LV-Details (Class starts on: 2024-09-27)
      Location: T9/SR 005 Übungsraum (Takustr. 9)
    • 19333002 Practice seminar
      Practice seminar for Cybersecurity and AI I (Gerhard Wunder)
      Schedule: Mo 14:00-16:00 (Class starts on: 2024-10-14)
      Location: T9/SR 006 Seminarraum (Takustr. 9)

  • Current Research Topics in Data Science Technologies

    0590bB2.4
    • 19327201 Lecture
      Data compression (Heiko Schwarz)
      Schedule: Mo 14:00-16:00 (Class starts on: 2024-10-14)
      Location: T9/049 Seminarraum (Takustr. 9)

      Comments

      Data compression is a technology, which only enables a variety of applications in our information age. Even though the underlying technology is often hidden from the end user, we use data compression every day when we hear music, watch images and videos, or use applications on our smartphone.

      In this course, the fundamental and most often used approaches for data compression are introduced.  We discuss theoretical foundations as well as methods used in practice.

      The first part of the course deals with lossless compression, in which the original data can be reconstructed exactly. This part includes the following topics:

      • Unique decodability and prefix codes
      • Entropy and entropy rate as theoretical limits of lossless compression
      • Optimal codes, Huffman codes
      • Arithmetic coding
      • Lempel-Ziv coding
      • Linear prediction
      • Examples from text, image and audio compression

      In the second part of the course, we consider lossy compression, by which only an approximation of the original data can be reconstructed. This type of compression enables much higher compression rates and is the dominant form of compression for audio, image and video data. The second part of the course includes the following topics:

      • Scalar quantization, optimal scalar quantization
      • Theoretical limits of lossy compression: Rate distortion functions
      • Vector quantization
      • Predictive quantization
      • Transform coding
      • Examples from audio, image, and video compression

      Suggested reading

      • Sayood, K. (2018), “Introduction to Data Compression,” Morgan Kaufmann, Cambridge, MA.
      • Cover, T. M. and Thomas, J. A. (2006), “Elements of Information Theory,” John Wiley & Sons, New York.
      • Gersho, A. and Gray, R. M. (1992), “Vector Quantization and Signal Compression,” Kluwer Academic Publishers, Boston, Dordrecht, London.
      • Jayant, N. S. and Noll, P. (1994), “Digital Coding of Waveforms,” Prentice-Hall, Englewood Cliffs, NJ, USA.
      • Wiegand, T. and Schwarz, H. (2010), “Source Coding: Part I of Fundamentals of Source and Video Coding,” Foundations and Trends in Signal Processing, vol. 4, no. 1-2.
    • 19328601 Lecture
      Cryptocurrencies and Blockchain (Katinka Wolter, Justus Purat)
      Schedule: Di 10:00-12:00 (Class starts on: 2024-10-15)
      Location: T9/049 Seminarraum (Takustr. 9)

      Comments

      We will study the history, technology and applications of cryptocurrencies and blockchain.

      Suggested reading

      Bitcoin and Cryptocurrency Technologies: A Comprehensive Introduction, by Arvind Narayanan, Joseph Bonneau, Edward Felten, Andrew Miller, Steven Goldfeder
    • 19333001 Lecture
      Cybersecurity and AI I: Privacy, Biometry, Certification (Gerhard Wunder)
      Schedule: Di 12:00-14:00 (Class starts on: 2024-10-15)
      Location: T9/SR 006 Seminarraum (Takustr. 9)
    • 19327202 Practice seminar
      Practice seminar for Data Compression (Heiko Schwarz)
      Schedule: Mo 12:00-14:00 (Class starts on: 2024-10-14)
      Location: T9/049 Seminarraum (Takustr. 9)
    • 19328602 Practice seminar
      Practice Session on Cryptocurrencies (Justus Purat)
      Schedule: Do 10:00-12:00, zusätzliche Termine siehe LV-Details (Class starts on: 2024-09-27)
      Location: T9/SR 005 Übungsraum (Takustr. 9)
    • 19333002 Practice seminar
      Practice seminar for Cybersecurity and AI I (Gerhard Wunder)
      Schedule: Mo 14:00-16:00 (Class starts on: 2024-10-14)
      Location: T9/SR 006 Seminarraum (Takustr. 9)

  • Selected Topics in Data Science Technologies (A)

    0590bB2.5
    • 19312101 Lecture
      Systems Software (Barry Linnert)
      Schedule: Mo 10:00-12:00, Do 12:00-14:00, zusätzliche Termine siehe LV-Details (Class starts on: 2024-10-14)
      Location: T9/049 Seminarraum (Takustr. 9)

      Additional information / Pre-requisites

      Language

      The course language is German as is the oral presentation of the lecturer, but the slides and all written material is available in English. You can always ask questions in English. The practice sheets and final exam are formulated in German, but may be answered in English, too.

      Homepage

      https://www.inf.fu-berlin.de/w/SE/VorlesungBetriebssysteme

      Comments

      Operating systems tie together the execution of applications, user experience and usability with the management of computer hardware. Starting with the tasks an operating system has to perform and the requirements it has to meet, the most important aspects of design and development of modern operating systems will be introduced:

      • Structure and design of an operating system including historical summary, structures and philosophies of OS design and resources and resource management
      • Threads and processes including thread management
      • Scheduling including real-time scheduling
      • Process interaction and inter-process communication
      • Resource management including device operation, driver development, management and operation of input- and output devices
      • Memory management including address spaces and virtual memory
      • File system including management and operation of discs and memory hierarchy
      • Distributed operating systems including distributed architectures for resource management
      • Performance evaluation and modeling including overload detection and handling

      Modern operating systems provide examples for different aspects and current research will be introduced. The tutorials serve to reflect the topics dealt with in the lecture and to acquire experience by developing a small operating system.

      Suggested reading

      • A.S. Tanenbaum: Modern Operating Systems, 2nd Ed. Prentice-Hall, 2001
      • A. Silberschatz et al.: Operating Systems Concepts with Java, 6th Ed. Wiley, 2004
    • 19315401 Lecture
      Graph and Network Algorithms (Günther Rothe)
      Schedule: Mo 14:00-16:00, Fr 12:00-14:00 (Class starts on: 2024-10-14)
      Location: T9/SR 005 Übungsraum (Takustr. 9)

      Additional information / Pre-requisites

      Target Audience

      Masters students in Computer Science or Mathematics, advanced Bachelor students.

      Prerequisites

      "Advanced Algorithms" or a similar class

      Comments

      Graphs and networks are an important modeling tool for all kinds of relations in Computer Science and beyond, for example social networks, traffic networks, and so on. We will treat algorithmic problems that arise in this context:

      • analysis of networks
      • optimization in graphs
      • graph drawing

      Suggested reading

      Wird noch bekannt gegeben.
    • 19312102 Practice seminar
      Practice seminar for Systems Software (Barry Linnert)
      Schedule: Mi 10:00-12:00 (Class starts on: 2024-10-16)
      Location: T9/049 Seminarraum (Takustr. 9)
    • 19315402 Practice seminar
      Practice seminar for Graph and Network Algorithms (Mahmoud Elashmawi, Günther Rothe)
      Schedule: Mi 12:00-14:00 (Class starts on: 2024-10-16)
      Location: T9/053 Seminarraum (Takustr. 9)

  • Selected Topics in Data Science Technologies (B)

    0590bB2.6
    • 19312101 Lecture
      Systems Software (Barry Linnert)
      Schedule: Mo 10:00-12:00, Do 12:00-14:00, zusätzliche Termine siehe LV-Details (Class starts on: 2024-10-14)
      Location: T9/049 Seminarraum (Takustr. 9)

      Additional information / Pre-requisites

      Language

      The course language is German as is the oral presentation of the lecturer, but the slides and all written material is available in English. You can always ask questions in English. The practice sheets and final exam are formulated in German, but may be answered in English, too.

      Homepage

      https://www.inf.fu-berlin.de/w/SE/VorlesungBetriebssysteme

      Comments

      Operating systems tie together the execution of applications, user experience and usability with the management of computer hardware. Starting with the tasks an operating system has to perform and the requirements it has to meet, the most important aspects of design and development of modern operating systems will be introduced:

      • Structure and design of an operating system including historical summary, structures and philosophies of OS design and resources and resource management
      • Threads and processes including thread management
      • Scheduling including real-time scheduling
      • Process interaction and inter-process communication
      • Resource management including device operation, driver development, management and operation of input- and output devices
      • Memory management including address spaces and virtual memory
      • File system including management and operation of discs and memory hierarchy
      • Distributed operating systems including distributed architectures for resource management
      • Performance evaluation and modeling including overload detection and handling

      Modern operating systems provide examples for different aspects and current research will be introduced. The tutorials serve to reflect the topics dealt with in the lecture and to acquire experience by developing a small operating system.

      Suggested reading

      • A.S. Tanenbaum: Modern Operating Systems, 2nd Ed. Prentice-Hall, 2001
      • A. Silberschatz et al.: Operating Systems Concepts with Java, 6th Ed. Wiley, 2004
    • 19315401 Lecture
      Graph and Network Algorithms (Günther Rothe)
      Schedule: Mo 14:00-16:00, Fr 12:00-14:00 (Class starts on: 2024-10-14)
      Location: T9/SR 005 Übungsraum (Takustr. 9)

      Additional information / Pre-requisites

      Target Audience

      Masters students in Computer Science or Mathematics, advanced Bachelor students.

      Prerequisites

      "Advanced Algorithms" or a similar class

      Comments

      Graphs and networks are an important modeling tool for all kinds of relations in Computer Science and beyond, for example social networks, traffic networks, and so on. We will treat algorithmic problems that arise in this context:

      • analysis of networks
      • optimization in graphs
      • graph drawing

      Suggested reading

      Wird noch bekannt gegeben.
    • 19312102 Practice seminar
      Practice seminar for Systems Software (Barry Linnert)
      Schedule: Mi 10:00-12:00 (Class starts on: 2024-10-16)
      Location: T9/049 Seminarraum (Takustr. 9)
    • 19315402 Practice seminar
      Practice seminar for Graph and Network Algorithms (Mahmoud Elashmawi, Günther Rothe)
      Schedule: Mi 12:00-14:00 (Class starts on: 2024-10-16)
      Location: T9/053 Seminarraum (Takustr. 9)

  • Seminar: Data Science Technology

    0590bB2.7
    • 19303811 Seminar
      Project Seminar: Data Management (Muhammed-Ugur Karagülle, Agnès Voisard)
      Schedule: Do 12:00-14:00 (Class starts on: 2024-10-17)
      Location: T9/137 Konferenzraum (Takustr. 9)

      Additional information / Pre-requisites

      Requirement

      ALP I-III, Foundations of Datenbase Systems, good programming knowledge.

      Comments

      Content

      A project seminar serves as preparation of a thesis (bachelor or master) in the AGDB. The focus of this project seminar lies on the analysis and visualization of medical data. Additionally, we will realize a small software project.

      Suggested reading

      Wird bekannt gegeben.
    • 19328217 Seminar / Undergraduate Course
      Seminar/Proseminar: New Trends in Information Systems (Agnès Voisard, Muhammed-Ugur Karagülle)
      Schedule: Mi 10:00-12:00 (Class starts on: 2024-10-16)
      Location: T9/SR 005 Übungsraum (Takustr. 9)

      Comments

      This seminar aims at studying recent trends in data management. Among others, we will look at two emerging topics, namely Location-Based Services (LBS) and Event-Based Services (EBS).

      Event-based Systems (EBS) are part of many current applications such as business activity monitoring, stock tickers, facility management, data streaming, or security. In the past years, the topic has gained increasing attention from both the industrial and the academic community. Current research concentrates of diverse aspects that range from event capture (incoming data) to response triggering. This seminar aims at studying some of the current trends in Event-based Systems with a strong focus on models and design. Location-based services are now often part of every day's life through applications such as navigation assistants in the public or private transportation domain. The underlying technology deals with many different aspects, such as location detection, information retrieval, or privacy. More recently, aspects such as user context and preferences were considered in order to send users more personalized information.

      A solid background in databases is required, typically a database course at a bachelor level.

      Suggested reading

      Wird bekannt gegeben.
    • 19333417 Seminar / Undergraduate Course
      Seminar/Proseminar: Explainable AI for Data Science (Grégoire Montavon)
      Schedule: Di 14:00-16:00 (Class starts on: 2024-10-15)
      Location: KöLu24-26/SR 006 Neuro/Mathe (Königin-Luise-Str. 24 / 26)

      Comments

      Explainable AI is a recent and growing subfield of machine learning (ML) that aims to bring transparency into ML models without sacrificing their predictive accuracy. This seminar will explore current research on the use of Explainable AI for extracting insights from large datasets of interest. Use cases in biomedicine, chemistry and earth sciences will be covered.

      Students will select a few papers from a pool of thematically relevant research papers, which they will read and present over the course of the semester.
    • 19334617 Seminar / Undergraduate Course
      Seminar/Proseminar: Multi-Agent Reinforcement Learning (Tim Landgraf)
      Schedule: Mo 10:00-12:00 (Class starts on: 2024-10-14)
      Location: A6/SR 007/008 Seminarraum (Arnimallee 6)

      Comments

      This seminar provides an exploration of large language models (LLMs), covering both foundational concepts and the latest advancements in the field. Participants will gain a comprehensive understanding of the architecture, training, and applications of LLMs, based on seminal research papers. The course will be organised as a journal club: students present individual papers, which are then discussed in the group to make sure we all get the ideas presented.

      ### Potential Topics

         - Neural networks and deep learning basics

         - Sequence modeling and RNNs (Recurrent Neural Networks)

         - Vaswani et al.'s "Attention is All You Need" paper

         - Self-attention mechanism

         - Multi-head attention and positional encoding

         - GPT-1: Radford et al.'s pioneering work

         - GPT-2: Scaling and implications

         - GPT-3: Architectural advancements and few-shot learning

         - BERT (Bidirectional Encoder Representations from Transformers)

         - T5 (Text-To-Text Transfer Transformer)

         - DistilBERT and efficiency improvements

         - Mamba:l and other SSMs: Design principles and performance

         - Flash Attention et al: Improving efficiency and scalability

         - Training regimes and resource requirements

         - Fine-tuning and transfer learning

      - Emergence of new capabilities
    • 19334717 Seminar / Undergraduate Course
      Seminar/Proseminar: Machine Learning for Process Control (Grégoire Montavon)
      Schedule: Do 14:00-16:00 (Class starts on: 2024-10-17)
      Location: KöLu24-26/SR 006 Neuro/Mathe (Königin-Luise-Str. 24 / 26)

      Comments

      Numerous real-world processes need to be kept under control in order to ensure safety or efficiency. Machine learning models are good candidates for this. They can for example detect shifts/anomalies/decalibrations/instabilities/etc. and possibly also predict which action needs to be taken on the process. The real-time nature of such tasks brings unique challenges from a ML perspective compared to classical application of ML. This seminar will explore relevant ML methods such as online/reinforcement learning and real-time data analysis. Use cases in manufacturing and intensive care will be covered. Students will select a few papers from a pool of thematically relevant research papers, which they will read and present over the course of the semester.
    • 19336311 Seminar
      Visualization for Artificial Intelligence Explainability (Georges Hattab)
      Schedule: Mi 10:00-12:00 (Class starts on: 2024-10-23)
      Location: A3/019 Seminarraum (Arnimallee 3-5)

      Comments

      As AI systems grow more powerful, there is an increasing need to make these complex "black box" models interpretable and explainable. This seminar explores how data visualization techniques can provide crucial insights into how AI models operate and arrive at their outputs. Cutting-edge methods like saliency maps, decision trees, and dimensionality reduction visualizations allow us to peer inside deep neural networks and understand what factors they are considering.

       

      The seminar also covers visualization literacy - effectively communicating AI explainability visualizations to different stakeholders. Case studies highlight best practices for visualizing model behavior, evaluating fairness, and instilling appropriate levels of trust. Attendees will gain an understanding of how visualization can demystify AI, foster transparency, and enable real-world deployment of these systems in high-stakes domains.
    • 19337211 Seminar
      Representation Learning (Georges Hattab)
      Schedule: Mi 14:00-16:00 (Class starts on: 2024-10-23)
      Location: A3/019 Seminarraum (Arnimallee 3-5)

      Comments

      While traditional feature engineering has been successful, modern machine learning increasingly relies on representation learning - automatically discovering informative features or representations from raw data. This seminar dives into advanced neural network-based approaches that learn dense vector representations capturing the underlying explanatory factors in complex, high-dimensional datasets.

       

      The seminar will cover techniques like autoencoders, variational autoencoders, and self-supervised contrastive learning methods that leverage unlabeled data to learn rich representations. You'll learn about properties of effective learned representations like preserving locality, handling sparse inputs, and disentangling underlying factors. Case studies demonstrate how representation learning enables breakthrough performance on tasks like image recognition and natural language understanding. You'll gain insights into interpreting these learned representations as well as their potential and limitations.

  • Software Project Data Science B

    0590bB2.8
    • 19308312 Project Seminar
      Implementation Project: Applications of Algorithms (László Kozma)
      Schedule: Di 14:00-16:00 (Class starts on: 2024-10-08)
      Location: T9/K 040 Multimediaraum (Takustr. 9)

      Comments

      Contents

      We choose a typical application area of algorithms, usually for geometric problems, and develop software solutions for it, e.g., computer graphics (representation of objects in a computer, projections, hidden edge and surface removal, lighting, raytracing), computer vision (image processing, filtering, projections, camera calibration, stereo-vision) or pattern recognition (classification, searching).

      Prerequsitions

      Basic knowledge in design and anaylsis of algorithms.

      Suggested reading

      je nach Anwendungsgebiet
    • 19315312 Project Seminar
      Software Project: Distributed Systems (Justus Purat)
      Schedule: Di 14:00-16:00 (Class starts on: 2024-10-15)
      Location: T9/053 Seminarraum (Takustr. 9)
    • 19334212 Project Seminar
      Softwareproject: Machine Learning with Graphs for Improved (Cancer) Treatment (Pauline Hiort, Pascal Iversen)
      Schedule: Mi 14:00-16:00 (Class starts on: 2024-10-16)
      Location: T9/K 040 Multimediaraum (Takustr. 9)

      Comments

      In the software project, we will implement, train, and evaluate various machine learning (ML) methods. The focus of the project is on graph neural networks (GNNs) that use graphs as input features for learning. We will compare the GNNs with various baseline methods, such as neural networks and regression models. The different ML methods will be applied and evaluated on a specific dataset, such as predicting drug combinations for cancer treatment. We will prepare the dataset and analyze it using the implemented methods. The programming language is Python, and we plan to use modern Python modules for ML like scikit-learn, TensorFlow, or PyTorch. Good Python skills are required. The goal is to create a Python package that provides reusable code for preprocessing, training ML models, and evaluating results with documentation (e.g., using Sphinx) for the specific use case. The software project takes place throughout the semester and can also be conducted in English.

  • Software Project Data Science A

    0590bB2.1
    • 19308312 Project Seminar
      Implementation Project: Applications of Algorithms (László Kozma)
      Schedule: Di 14:00-16:00 (Class starts on: 2024-10-08)
      Location: T9/K 040 Multimediaraum (Takustr. 9)

      Comments

      Contents

      We choose a typical application area of algorithms, usually for geometric problems, and develop software solutions for it, e.g., computer graphics (representation of objects in a computer, projections, hidden edge and surface removal, lighting, raytracing), computer vision (image processing, filtering, projections, camera calibration, stereo-vision) or pattern recognition (classification, searching).

      Prerequsitions

      Basic knowledge in design and anaylsis of algorithms.

      Suggested reading

      je nach Anwendungsgebiet
    • 19314012 Project Seminar
      Software Project: Semantic Technologies (Adrian Paschke)
      Schedule: Mi 14:00-16:00 (Class starts on: 2024-10-16)
      Location: A7/SR 031 (Arnimallee 7)

      Additional information / Pre-requisites

      Further information can be found on the course website of the AG Corporate Semantic Web.

      Comments

      Mixed groups of master and bachelor students will either implement an independent project or are part of a larger project in the area of semantic AI technologies. They will gain in-depth programming knowledge about applications of semantic technologies and artificial intelligence techniques in the Corporate Semantic Web. They will practice teamwork and best practices in software development of AI systems and Semantic Web applications. The software project can be done in collaboration with an external partner from industry or standardization. It is possible to continue the project as bachelor or master thesis.
    • 19315312 Project Seminar
      Software Project: Distributed Systems (Justus Purat)
      Schedule: Di 14:00-16:00 (Class starts on: 2024-10-15)
      Location: T9/053 Seminarraum (Takustr. 9)
    • 19334212 Project Seminar
      Softwareproject: Machine Learning with Graphs for Improved (Cancer) Treatment (Pauline Hiort, Pascal Iversen)
      Schedule: Mi 14:00-16:00 (Class starts on: 2024-10-16)
      Location: T9/K 040 Multimediaraum (Takustr. 9)

      Comments

      In the software project, we will implement, train, and evaluate various machine learning (ML) methods. The focus of the project is on graph neural networks (GNNs) that use graphs as input features for learning. We will compare the GNNs with various baseline methods, such as neural networks and regression models. The different ML methods will be applied and evaluated on a specific dataset, such as predicting drug combinations for cancer treatment. We will prepare the dataset and analyze it using the implemented methods. The programming language is Python, and we plan to use modern Python modules for ML like scikit-learn, TensorFlow, or PyTorch. Good Python skills are required. The goal is to create a Python package that provides reusable code for preprocessing, training ML models, and evaluating results with documentation (e.g., using Sphinx) for the specific use case. The software project takes place throughout the semester and can also be conducted in English.

  • Modules w/o courses in this semester

    16 Modules
    • Pattern Recognition 0089cA1.12
    • Network-Based Information Systems 0089cA1.13
    • Computer Security 0089cA1.16
    • Distributed Systems 0089cA1.20
    • Artificial Intelligence 0089cA1.9
    • Mobile Communications 0089cA3.3
    • Applied Machine Learning in Bioinformatics 0262cD1.12
    • Machine Learning in Bioinformatics 0262cD1.7
    • Big Data Analysis in Bioinformatics 0262cD1.8
    • Data Science in the Life Sciences 0590bB1.1
    • Research Practice 0590bB1.2
    • Interdisciplinary Apporaches (Data Science) A 0590bB1.27
    • Interdisciplinary Apporaches (Data Science) B 0590bB1.28
    • Ethical Foundations of Data Science 0590bB1.3
    • Data Base Systems for students of Data Science 0590bB2.9
    • Accompanying colloquium 0590bE1.2