Search result: Catalogue data in Spring Semester 2021
|Computer Science Master|
|Master Studies (Programme Regulations 2020)|
|Major in Machine Intelligence|
|261-5110-00L||Optimization for Data Science||W||10 credits||3V + 2U + 4A||B. Gärtner, D. Steurer, N. He|
|Abstract||This course provides an in-depth theoretical treatment of optimization methods that are particularly relevant in data science.|
|Objective||Understanding the theoretical guarantees (and their limits) of relevant optimization methods used in data science. Learning general paradigms to deal with optimization problems arising in data science.|
|Content||This course provides an in-depth theoretical treatment of optimization methods that are particularly relevant in machine learning and data science.|
In the first part of the course, we will first give a brief introduction to convex optimization, with some basic motivating examples from machine learning. Then we will analyse classical and more recent first and second order methods for convex optimization: gradient descent, Nesterov's accelerated method, proximal and splitting algorithms, subgradient descent, stochastic gradient descent, variance-reduced methods, Newton's method, and Quasi-Newton methods. The emphasis will be on analysis techniques that occur repeatedly in convergence analyses for various classes of convex functions. We will also discuss some classical and recent theoretical results for nonconvex optimization.
In the second part, we discuss convex programming relaxations as a powerful and versatile paradigm for designing efficient algorithms to solve computational problems arising in data science. We will learn about this paradigm and develop a unified perspective on it through the lens of the sum-of-squares semidefinite programming hierarchy. As applications, we are discussing non-negative matrix factorization, compressed sensing and sparse linear regression, matrix completion and phase retrieval, as well as robust estimation.
|Prerequisites / Notice||As background, we require material taught in the course "252-0209-00L Algorithms, Probability, and Computing". It is not necessary that participants have actually taken the course, but they should be prepared to catch up if necessary.|
|263-3710-00L||Machine Perception |
Number of participants limited to 200.
|W||8 credits||3V + 2U + 2A||O. Hilliges, S. Tang|
|Abstract||Recent developments in neural networks (aka “deep learning”) have drastically advanced the performance of machine perception systems in a variety of areas including computer vision, robotics, and intelligent UIs. This course is a deep dive into deep learning algorithms and architectures with applications to a variety of perceptual tasks.|
|Objective||Students will learn about fundamental aspects of modern deep learning approaches for perception. Students will learn to implement, train and debug their own neural networks and gain a detailed understanding of cutting-edge research in learning-based computer vision, robotics and HCI. The final project assignment will involve training a complex neural network architecture and applying it on a real-world dataset of human activity.|
The core competency acquired through this course is a solid foundation in deep-learning algorithms to process and interpret human input into computing systems. In particular, students should be able to develop systems that deal with the problem of recognizing people in images, detecting and describing body parts, inferring their spatial configuration, performing action/gesture recognition from still images or image sequences, also considering multi-modal data, among others.
|Content||We will focus on teaching: how to set up the problem of machine perception, the learning algorithms, network architectures and advanced deep learning concepts in particular probabilistic deep learning models |
The course covers the following main areas:
I) Foundations of deep-learning.
II) Probabilistic deep-learning for generative modelling of data (latent variable models, generative adversarial networks and auto-regressive models).
III) Deep learning in computer vision, human-computer interaction and robotics.
Specific topics include:
I) Deep learning basics:
a) Neural Networks and training (i.e., backpropagation)
b) Feedforward Networks
c) Timeseries modelling (RNN, GRU, LSTM)
d) Convolutional Neural Networks for classification
II) Probabilistic Deep Learning:
a) Latent variable models (VAEs)
b) Generative adversarial networks (GANs)
c) Autoregressive models (PixelCNN, PixelRNN, TCNs)
III) Deep Learning techniques for machine perception:
a) Fully Convolutional architectures for dense per-pixel tasks (i.e., instance segmentation)
b) Pose estimation and other tasks involving human activity
c) Deep reinforcement learning
IV) Case studies from research in computer vision, HCI, robotics and signal processing
Book by Ian Goodfellow and Yoshua Bengio
|Prerequisites / Notice||*** |
In accordance with the ETH Covid-19 master plan the lecture will be fully virtual. Details on the course website.
This is an advanced grad-level course that requires a background in machine learning. Students are expected to have a solid mathematical foundation, in particular in linear algebra, multivariate calculus, and probability. The course will focus on state-of-the-art research in deep-learning and will not repeat basics of machine learning
Please take note of the following conditions:
1) The number of participants is limited to 200 students (MSc and PhDs).
2) Students must have taken the exam in Machine Learning (252-0535-00) or have acquired equivalent knowledge
3) All practical exercises will require basic knowledge of Python and will use libraries such as Pytorch, scikit-learn and scikit-image. We will provide introductions to Pytorch and other libraries that are needed but will not provide introductions to basic programming or Python.
The following courses are strongly recommended as prerequisite:
* "Visual Computing" or "Computer Vision"
The course will be assessed by a final written examination in English. No course materials or electronic devices can be used during the examination. Note that the examination will be based on the contents of the lectures, the associated reading materials and the exercises.
|252-0526-00L||Statistical Learning Theory||W||8 credits||3V + 2U + 2A||J. M. Buhmann, C. Cotrini Jimenez|
|Abstract||The course covers advanced methods of statistical learning: |
- Variational methods and optimization.
- Deterministic annealing.
- Clustering for diverse types of data.
- Model validation by information theory.
|Objective||The course surveys recent methods of statistical learning. The fundamentals of machine learning, as presented in the courses "Introduction to Machine Learning" and "Advanced Machine Learning", are expanded from the perspective of statistical learning.|
|Content||- Variational methods and optimization. We consider optimization approaches for problems where the optimizer is a probability distribution. We will discuss concepts like maximum entropy, information bottleneck, and deterministic annealing.|
- Clustering. This is the problem of sorting data into groups without using training samples. We discuss alternative notions of "similarity" between data points and adequate optimization procedures.
- Model selection and validation. This refers to the question of how complex the chosen model should be. In particular, we present an information theoretic approach for model validation.
- Statistical physics models. We discuss approaches for approximately optimizing large systems, which originate in statistical physics (free energy minimization applied to spin glasses and other models). We also study sampling methods based on these models.
|Lecture notes||A draft of a script will be provided. Lecture slides will be made available.|
|Literature||Hastie, Tibshirani, Friedman: The Elements of Statistical Learning, Springer, 2001.|
L. Devroye, L. Gyorfi, and G. Lugosi: A probabilistic theory of pattern recognition. Springer, New York, 1996
|Prerequisites / Notice||Knowledge of machine learning (introduction to machine learning and/or advanced machine learning)|
Basic knowledge of statistics.
|252-0579-00L||3D Vision||W||5 credits||3G + 1A||M. Pollefeys, V. Larsson|
|Abstract||The course covers camera models and calibration, feature tracking and matching, camera motion estimation via simultaneous localization and mapping (SLAM) and visual odometry (VO), epipolar and mult-view geometry, structure-from-motion, (multi-view) stereo, augmented reality, and image-based (re-)localization.|
|Objective||After attending this course, students will:|
1. understand the core concepts for recovering 3D shape of objects and scenes from images and video.
2. be able to implement basic systems for vision-based robotics and simple virtual/augmented reality applications.
3. have a good overview over the current state-of-the art in 3D vision.
4. be able to critically analyze and asses current research in this area.
|Content||The goal of this course is to teach the core techniques required for robotic and augmented reality applications: How to determine the motion of a camera and how to estimate the absolute position and orientation of a camera in the real world. This course will introduce the basic concepts of 3D Vision in the form of short lectures, followed by student presentations discussing the current state-of-the-art. The main focus of this course are student projects on 3D Vision topics, with an emphasis on robotic vision and virtual and augmented reality applications.|
|252-3005-00L||Natural Language Processing |
Number of participants limited to 400.
|W||5 credits||2V + 1U + 1A||R. Cotterell|
|Abstract||This course presents topics in natural language processing with an emphasis on modern techniques, primarily focusing on statistical and deep learning approaches. The course provides an overview of the primary areas of research in language processing as well as a detailed exploration of the models and techniques used both in research and in commercial natural language systems.|
|Objective||The objective of the course is to learn the basic concepts in the statistical processing of natural languages. The course will be project-oriented so that the students can also gain hands-on experience with state-of-the-art tools and techniques.|
|Content||This course presents an introduction to general topics and techniques used in natural language processing today, primarily focusing on statistical approaches. The course provides an overview of the primary areas of research in language processing as well as a detailed exploration of the models and techniques used both in research and in commercial natural language systems.|
|Literature||Jacob Eisenstein: Introduction to Natural Language Processing (Adaptive Computation and Machine Learning series)|
|261-5120-00L||Machine Learning for Health Care |
Number of participants limited to 150.
|W||5 credits||3P + 1A||V. Boeva, G. Rätsch, J. Vogt|
|Abstract||The course will review the most relevant methods and applications of Machine Learning in Biomedicine, discuss the main challenges they present and their current technical problems.|
|Objective||During the last years, we have observed a rapid growth in the field of Machine Learning (ML), mainly due to improvements in ML algorithms, the increase of data availability and a reduction in computing costs. This growth is having a profound impact in biomedical applications, where the great variety of tasks and data types enables us to get benefit of ML algorithms in many different ways. In this course we will review the most relevant methods and applications of ML in biomedicine, discuss the main challenges they present and their current technical solutions.|
|Content||The course will consist of four topic clusters that will cover the most relevant applications of ML in Biomedicine: |
1) Structured time series: Temporal time series of structured data often appear in biomedical datasets, presenting challenges as containing variables with different periodicities, being conditioned by static data, etc.
2) Medical notes: Vast amount of medical observations are stored in the form of free text, we will analyze stategies for extracting knowledge from them.
3) Medical images: Images are a fundamental piece of information in many medical disciplines. We will study how to train ML algorithms with them.
4) Genomics data: ML in genomics is still an emerging subfield, but given that genomics data are arguably the most extensive and complex datasets that can be found in biomedicine, it is expected that many relevant ML applications will arise in the near future. We will review and discuss current applications and challenges.
|Prerequisites / Notice||Data Structures & Algorithms, Introduction to Machine Learning, Statistics/Probability, Programming in Python, Unix Command Line|
Relation to Course 261-5100-00 Computational Biomedicine: This course is a continuation of the previous course with new topics related to medical data and machine learning. The format of Computational Biomedicine II will also be different. It is helpful but not essential to attend Computational Biomedicine before attending Computational Biomedicine II.
|263-5000-00L||Computational Semantics for Natural Language Processing |
Limited number of participants: 80. Last cancellation/deregistration date for this graded semester performance: Friday, 26 March 2021! Please note that after that date no deregistration will be accepted and the course will be considered as "fail".
|W||6 credits||2V + 1U + 2A||M. Sachan|
|Abstract||This course presents an introduction to Natural language processing (NLP) with an emphasis on computational semantics i.e. the process of constructing and reasoning with meaning representations of natural language text.|
|Objective||The objective of the course is to learn about various topics in computational semantics and its importance in natural language processing methodology and research. Exercises and the project will be key parts of the course so the students will be able to gain hands-on experience with state-of-the-art techniques in the field.|
|Content||We will take a modern view of the topic, and focus on various statistical and deep learning approaches for computation semantics. We will also overview various primary areas of research in language processing and discuss how the computational semantics view can help us make advances in NLP.|
|Lecture notes||Lecture slides will be made available at the course Web site.|
|Literature||No textbook is required, but there will be regularly assigned readings from research literature, linked to the course website.|
|Prerequisites / Notice||The student should have successfully completed a graduate level class in machine learning (252-0220-00L), deep learning (263-3210-00L) or natural language processing (252-3005-00L) before. Similar courses from other universities are acceptable too.|
|263-5300-00L||Guarantees for Machine Learning |
Number of participants limited to 30.
Last cancellation/deregistration date for this graded semester performance: 17 March 2021! Please note that after that date no deregistration will be accepted and a "no show" will appear on your transcript.
|W||7 credits||3G + 3A||F. Yang|
|Abstract||This course is aimed at advanced master and doctorate students who want to conduct independent research on theory for modern machine learning (ML). It teaches classical and recent methods in statistical learning theory commonly used to prove theoretical guarantees for ML algorithms. The knowledge is then applied in independent project work that focuses on understanding modern ML phenomena.|
|Objective||Learning objectives: |
- acquire enough mathematical background to understand a good fraction of theory papers published in the typical ML venues. For this purpose, students will learn common mathematical techniques from statistics and optimization in the first part of the course and apply this knowledge in the project work
- critically examine recently published work in terms of relevance and determine impactful (novel) research problems. This will be an integral part of the project work and involves experimental as well as theoretical questions
- find and outline an approach (some subproblem) to prove a conjectured theorem. This will be practiced in lectures / exercise and homeworks and potentially in the final project.
- effectively communicate and present the problem motivation, new insights and results to a technical audience. This will be primarily learned via the final presentation and report as well as during peer-grading of peer talks.
|Content||This course touches upon foundational methods in statistical learning theory aimed at proving theoretical guarantees for machine learning algorithms, touching on the following topics|
- concentration bounds
- uniform convergence and empirical process theory
- high-dimensional statistics (e.g. sparsity)
- regularization for non-parametric statistics (e.g. in RKHS, neural networks)
- implicit regularization via gradient descent (e.g. margins, early stopping)
- minimax lower bounds
The project work focuses on current theoretical ML research that aims to understand modern phenomena in machine learning, including but not limited to
- how overparameterization could help generalization ( RKHS, NN )
- how overparameterization could help optimization ( non-convex optimization, loss landscape )
- complexity measures and approximation theoretic properties of randomly initialized and trained NN
- generalization of robust learning ( adversarial robustness, standard and robust error tradeoff, distribution shift)
|Prerequisites / Notice||It’s absolutely necessary for students to have a strong mathematical background (basic real analysis, probability theory, linear algebra) and good knowledge of core concepts in machine learning taught in courses such as “Introduction to Machine Learning”, “Regression”/ “Statistical Modelling”. In addition to these prerequisites, this class requires a high degree of mathematical maturity—including abstract thinking and the ability to understand and write proofs. |
Students have usually taken a subset of Fundamentals of Mathematical Statistics, Probabilistic AI, Neural Network Theory, Optimization for Data Science, Advanced ML, Statistical Learning Theory, Probability Theory (D-MATH)
|227-0434-10L||Mathematics of Information||W||8 credits||3V + 2U + 2A||H. Bölcskei|
|Abstract||The class focuses on mathematical aspects of |
1. Information science: Sampling theorems, frame theory, compressed sensing, sparsity, super-resolution, spectrum-blind sampling, subspace algorithms, dimensionality reduction
2. Learning theory: Approximation theory, greedy algorithms, uniform laws of large numbers, Rademacher complexity, Vapnik-Chervonenkis dimension
|Objective||The aim of the class is to familiarize the students with the most commonly used mathematical theories in data science, high-dimensional data analysis, and learning theory. The class consists of the lecture, exercise sessions with homework problems, and of a research project, which can be carried out either individually or in groups. The research project consists of either 1. software development for the solution of a practical signal processing or machine learning problem or 2. the analysis of a research paper or 3. a theoretical research problem of suitable complexity. Students are welcome to propose their own project at the beginning of the semester. The outcomes of all projects have to be presented to the entire class at the end of the semester.|
|Content||Mathematics of Information|
1. Signal representations: Frame theory, wavelets, Gabor expansions, sampling theorems, density theorems
2. Sparsity and compressed sensing: Sparse linear models, uncertainty relations in sparse signal recovery, super-resolution, spectrum-blind sampling, subspace algorithms (ESPRIT), estimation in the high-dimensional noisy case, Lasso
3. Dimensionality reduction: Random projections, the Johnson-Lindenstrauss Lemma
Mathematics of Learning
4. Approximation theory: Nonlinear approximation theory, best M-term approximation, greedy algorithms, fundamental limits on compressibility of signal classes, Kolmogorov-Tikhomirov epsilon-entropy of signal classes, optimal compression of signal classes
5. Uniform laws of large numbers: Rademacher complexity, Vapnik-Chervonenkis dimension, classes with polynomial discrimination
|Lecture notes||Detailed lecture notes will be provided at the beginning of the semester.|
|Prerequisites / Notice||This course is aimed at students with a background in basic linear algebra, analysis, statistics, and probability. |
We encourage students who are interested in mathematical data science to take both this course and "401-4944-20L Mathematics of Data Science" by Prof. A. Bandeira. The two courses are designed to be complementary.
H. Bölcskei and A. Bandeira
|401-3632-00L||Computational Statistics||W||8 credits||3V + 1U||M. Mächler|
|Abstract||We discuss modern statistical methods for data analysis, including methods for data exploration, prediction and inference. We pay attention to algorithmic aspects, theoretical properties and practical considerations. The class is hands-on and methods are applied using the statistical programming language R.|
|Objective||The student obtains an overview of modern statistical methods for data analysis, including their algorithmic aspects and theoretical properties. The methods are applied using the statistical programming language R.|
|Content||See the class website|
|Prerequisites / Notice||At least one semester of (basic) probability and statistics.|
Programming experience is helpful but not required.
- Page 1 of 1