Vorlesung: Statistical Machine Learning
|Teaching Assistants:||Daniel Palenicek, An Thai Le, Theo Gruner, MaximilianTölle, Theo Vincent|
|Lectures:||Monday, 11:40-13:20 in Room S103/226|
|Office Hours:||Thursday, 13:00 - 14:00 in Room S2|02, A126|
|Or request by Email: firstname.lastname@example.org|
|TU-CAN:||20-00-0358-iv Statistisches Maschinelles Lernen|
|Credits:||6,0 ECTS (4 SWS)|
Welcome to the Statistical Machine Learning (SML) lecture!
In today's world, data is being generated at an unprecedented rate. Billions of web pages are at our disposal, videos with an accumulated time of 500 hours are uploaded every minute on Youtube and the supermarket chain Walmart alone performed more than one million transactions per hour, creating a database of more than 2.5 petabytes of information. John Naisbitt has stated the problem very clearly: "We are drowning in information and starving for knowledge."
That's where Statistical Machine Learning comes in. It's a subfield of artificial intelligence that combines statistical techniques and algorithms to build models that can learn from data and make predictions or decisions based on that learning. Therefore, it helps us unlock valuable insights from the randomness in our world and lets us better understand it.
For instance, in healthcare, Statistical Machine Learning models can analyze large amounts of patient data to identify patterns and predict potential health risks. This can lead to earlier diagnosis and more effective treatment, ultimately improving patient outcomes. Similarly, in finance, SML models can analyze financial data to identify fraud, predict market trends, and develop more accurate risk models. This can help financial institutions make better investment decisions and minimize risk. In transportation, Statistical Machine Learning can help optimize traffic flow and reduce congestion by analyzing data from traffic sensors and GPS devices. This can ultimately lead to more efficient transportation systems and reduced carbon emissions.
To sum it up, whether you are interested in building predictive models, understanding data better, or simply curious about the intersection of statistics and artificial intelligence, this introductory lecture is for you!
The course gives an introduction to statistical machine learning methods, from foundational to recent advanced approaches. The following topics are expected to be covered throughout the semester:
- Probability Distributions
- Linear Models for Regression and Classification
- Kernel Methods, Graphical Models
- Mixture Models and EM
- Gaussian Processes
- Approximate Inference
- Latent Variable Representation
- Neural Networks
Math classes from the bachelor's degree, basic programming abilities, introductory classes to computer science.
The final exam date will be announced during the semester. It will cover all material presented in the lectures, unless specified otherwise. It will consist of roughly 30 questions and will take 90 minutes. Students are allowed to bring a cheat sheet consisting of a single A4 paper to the exam. The paper must be handwritten (not printed) and you can write on both faces. Students are allowed to use a non-electronic dictionary and a non-programmable calculator. For the final exam you can gain bonuspoints by successfully submitting the homework. Throughout the semester, we will ask you to solve four homeworks which contain theoretical questions and programming exercises.
All course announcements go through moodle -- make sure to subscribe to the forums to not miss homework releases and important announcements
Jan Peters heads the Intelligent Autonomous Systems Lab at the Department of Computer Science at the TU Darmstadt. Jan has studied computer science, electrical, control, mechanical and aerospace engineering. You can find Jan Peters in the Robert-Piloty building S2 | 02 room E314. You can also contact him through
Daniel Palenicek joined the Intelligent Autonomous System lab on October 1st, 2021 as a Ph.D. student. He is part of the 3AI project with Hessian.AI.
In his research, Daniel focuses on increasing sample efficiency of model-based reinforcement learning algorithms by studying the impact which model-errors have on the learning. You can contact him via email
An Thai Le joined the Intelligent Autonomous System lab on November 1st, 2021, as a Ph.D. student. He is currently working on applying Optimal Transport methods for the correspondence problem of Imitation Learning and also Motion Planning problems. He aims to explore fascinating viewpoints from Computational Geometry and apply these perspectives to Robotics and Machine Learning. He can be reached via email
Theo Vincent joined the Intelligent Autonomous System lab on December 1st, 2022, as a Ph.D. student. He is currently doing some research on Q learning algorithms, focusing on their exploiting behavior. He can be reached via email
Theo Gruner joined the Intelligent Autonomous System lab on September 15, 2022, as a Ph.D. student. He is part of the 3AI project with Hessian.AI. Theo is researching Bayesian inference methods for system identification of black-box simulators, focusing on their broad applicability and scalability. You can reach him through
Maximilian Tölle joined the Intelligent Autonomous System lab on November 1st, 2022, as a Ph.D. student. He is part of the research department Systems AI for Robot Learning of DFKI. Max is doing research on language interfaces for robot learning to improve human-robot interaction and ultimately include foundation models into the learning process. You can reach him through
For further inquiries do not hesitate to contact us immediately!