Machine learning is a field of scientific study concerned with algorithmic techniques that enable machines to learn performance on a given task via the discovery of patterns or regularities in exemplary data. Consequently, its methods commonly draw upon a statistical basis in conjunction with the computational capabilities of modern computing hardware.

This course aims to acquaint the student with the main branches of machine learning and provide a thorough introduction to the most widely used approaches and methods in this field.

1. Introduction to Machine Learning
   1. Regression & Classification
   2. Supervised & Unsupervised Learning
   3. Reinforcement Learning
2. Clustering
   1. Introduction to clustering
   2. K-Means
   3. Expectation Maximization
   4. DBScan
   5. Hierarchical Clustering
3. Regression
   1. Linear & Non-linear Regression
   2. Logistic Regression
   3. Quantile Regression
   4. Multivariate Regression
   5. Lasso & Ridge Regression
4. Support Vector Machines
   1. Introduction to Support Vector Machines
   2. SVM for Classification
   3. SVM for Regression
5. Decision Trees
   1. Introduction to Decision Trees
   2. Decision Trees for Classification
   3. Decision Trees for Regression
6. Genetic Algorithms
   1. Introduction to Genetic Algorithms
   2. Applications of Genetic Algorithms

Neural networks and deep learning approaches have revolutionized the fields of data science and artificial intelligence in recent years, and applications built on these techniques have reached or surpassed human performance in many specialized applications.

After a short review of the origins of neural networks and deep learning, this course will cover the most common neural network architectures and discuss in detail how neural networks are trained using dedicated data samples, avoiding common pitfalls such as overtraining.

The course includes a detailed overview of alternative methods to train neural networks and further network architectures which are relevant in a wide range of specialized application scenarios.

1. Introduction to Neural Network and Deep Learning
   1. The Biological Brain
   2. Perceptron and Multi-Layer Perceptrons
2. Network Architectures
   1. Feed-Forward Networks
   2. Convolutional Networks
   3. Recurrent Networks, Memory Cells and LSTMs
3. Neural Network Training
   1. Weight Initialization and Transfer Function
   2. Backpropagation and Gradient Descent
   3. Regularization and Overtraining
4. Alternative Training Methods
   1. Attention
   2. Feedback Alignment
   3. Synthetic Gradients
   4. Decoupled Network Interfaces
5. Further Network Architectures
   1. Generative Adversarial Networks
   2. Autoencoders
   3. Restricted Boltzmann Machines
   4. Capsule Networks
   5. Spiking Networks

Reinforcement learning allows computers to derive problem-solving strategies without being explicitly programmed for the specific task, similar to the way humans and animals learn.

After introducing the concepts of reinforcement learning, the course discusses the properties of Markov chains and single- and multi-armed bandits in detail. Special attention is given to the understanding of value functions and discounted value functions.

The course connects reinforcement learning with neural networks and deep learning and discusses how Q-Learning approaches can be used to utilize deep learning methods in reinforcement learning problems, including extensions such as double Q-Learning, hierarchical learning, and actor-critic learning.

Finally, the course discusses reinforcement learning approaches such as model-free and model-based learning and the tradeoff between exploration and exploitation.

1. Introduction to Reinforcement Learning
   1. Understanding Reinforcement Learning
   2. Components of Reinforcement Learning Systems
2. Markov Chains
   1. Markov Decision Process & Markov Property
   2. Value Functions and Discounted Value Functions
   3. General Utility Function
   4. Actions & Policy
   5. Bellman’s Equation
   6. Value Iteration
   7. Markov Chain Monte Carlo (MCMC)
3. Bandit
   1. Single-Arm Bandit
   2. Multi-Arm Bandit
4. Q-Learning
   1. Time-difference Learning
   2. Reinforcement Learning with Neural Networks & Deep Q Learning
   3. Experience Replay
   4. Double Q-Learning
   5. Delayed Sparse Rewards
   6. Hierarchical Learning
   7. Value- vs Policy-Based Learning
   8. Actor Critic Learning
5. Reinforcement Learning Approaches
   1. Model-Free Learning
   2. Model-Based Learning
   3. Exploration vs Exploitation

The topic of artificial Intelligence (AI) has been addressed in computer science and cognitive science research since the 1950s; however, the meaning associated with the term has changed considerably over time. Having once been predominantly associated with logical calculus, reasoning, and planning, AI is now primarily interpreted in the context of deep networks of computational units. Despite these changes in approach, the important characteristic of AI continues to be the understanding and reproduction of cognitive abilities and functions by machines. This seminar strives to elucidate current research trends in AI. The students learn to independently analyze selected topics and case studies and link them with well-known concepts, as well as critically question and discuss them.

• The seminar covers current topics in artificial intelligence. Each participant must write a seminar paper on a topic assigned to him/her.