Study Autonomous Learning

Autonomous Learning research aims at understanding how autonomous systems can efficiently learn from the interaction with the environment, especially by having an integrated approach to decision making and learning, allowing systems to autonomously decide on actions, representations, hyperparameters and model structures for the purpose of efficient learning.

We hope to foster the enthusiasm of students for this exciting research area and listed some of the many lectures which are offered by several universities and institutes on Autonomous Learning. For more information or scripts see the connected websites below.

In September 2014, we supported a Summer School on Autonomous Learning at Max Planck Institute for Mathematics in the Sciences in Leipzig. You may view all tutorials at the Summer School’s website. There will be a second Summer School supported by this Priority Programme in 2017.

TitleLecturerAbstract
Introduction to RoboticsMarc Toussaint (Stuttgart)

Robotics is an ultimate test of our progress in Artificial Intelligence, Machine Learning and Control Theory research. However, while these research fields consider general but idealized problem formulations, robotics has to deal with the specifics our concrete 3-dimensional physical world and eventually integrate methods and hardware in autonomous systems. Therefore robotics is more than an application of the above fields and requires specific knowledge of how to generate motion, physically interact with the environment and perceive it.

The lecture will give an introduction to robotics in four chapters:

Kinematics & Dynamics
Planning and optimization
Control Theory
Mobile robots
Vertiefung Neuronale NetzeRobert Haschke (Bielfeld)

Mit dieser Vorlesung werden Grundkenntnisse im Bereich neuronaler Netze und Lernalgorithmen vertieft. Behandelt werden weitere Netzmodelle wie Lokal Lineare Karten, Hyperbolische Selbstorganisierenden Karten, Growing Neural Gas, Radiale Basisfunktionen, sowie Eigenschaften und Lernverfahren für dynamische rekurrente Netze, insbesondere zur Zeitserienvorhersage. Dabei werden häufig Beispiele aus der Mustererkennung herangezogen, um praktische Aspekte wie Vorverarbeitung, Merkmalselektion, Techniken zur Konvergenzbeschleunigung und Wahl einer geeigneten Netzarchitektur zu illustrieren.

Maschinelles Lernen 1 - GrundverfahrenRüdiger Dillmann (Karlsruhe)

Die Vorlesung behandelt sowohl symbolische Lernverfahren, wie induktives Lernen (Lernen aus Beispielen, Lernen durch Beobachtung), deduktives Lernen (Erklärungsbasiertes Lernen) und Lernen aus Analogien, als auch subsymbolische Techniken wie Neuronale Netze, Support Vektor-Maschinen und Genetische Algorithmen.

Maschinelles Lernen 2 - Fortgeschrittene VerfahrenRüdiger Dillmann (Karlsruhe)

Der Schwerpunkt dieser Vorlesung liegt in der Einbettung und Anwendung von maschinell lernenden Verfahren in Entscheidungs- und Inferenzsystemen beginnend bei Methoden der Dimensionsreduktion, Merkmalsselektion/-bewertung über semi-überwachtes Lernen (semi-supervised learning) hin zu Methoden der probabilistischen Inferenz (wie z.B. Dempster Shafer Informationsfusion, Dynamischen und objektorientierte Bayessche Netze, POMDP, etc). Die Vorlesung führt in die Grundprinzipien sowie Grundstrukturen ein und erläutert bisher entwickelte Algorithmen.

 

Statistische Lernverfahren und MustererkennungReinhold Häb-Umbach (Paderborn)

The course on Statistical Learning and Pattern Recognition presents an introduction into the components and algorithms prevalent in statistical pattern recognition. Both parametric and non-parametric density estimation and classification techniques will be presented, as well as supervised and unsupervised learning paradigms. The presented techniques can be applied to a variety of classification problems, both for one-dimensional input data (e.g., speech), two-dimensional (e.g., image) or symbolic input data (e.g., documents).


 

Maschinelles LernenKlaus-Robert Müller (Berlin)

In dieser Vorlesungen werden weiterführende Themen des Maschinellen Lernens behandelt. Ein besonderer Schwerpunkt wird hierbei auf Anwendungen gelegt. Mehrere erfolgreiche Anwendungen des Maschinellen Lernens werden besprochen und auf die jeweiligen Besonderheiten wird eingegangen. Unter anderem werden folgende Themen behandelt:

  • Dimensionsreduktion
  • Blind-Source-Separation
  • Deep Learning
  • Kernmethoden für strukturierte Daten
  • Multiple-Kernel Learning
  • Optimierungstheorie
Machine Intelligence II (unsupervised methods)Klaus Obermayer (Berlin)

Topics covered

  • Probabilities and densities
  • Density estimation
  • Maximum likelihood
  • Principal Component Analysis
  • Hebbian learning
  • Kernel PCA
  • Independent Component Analysis
  • Stochastic optimization
  • K-means clustering
  • Pairwise clustering
  • Self-Organizing Maps
Autonomous Robotics: Action, Perception, and CognitionGregor Schöner (Bochum)

Neuroinformatics is concerned with the discovery of new solutions to technical problems of information processing. These solutions are sought based on analogies with nervous systems and the behavior of organisms. This course focuses on three exemplary problems to illustrate this approach:

  1. Artificial action (autonomous robotics)
  2. Artificial perception (robot vision)
  3. Artificial cognition (simplest cognitive capabilities of autonomous robots such as decision making, memory, behavioral organization).

The main methodological emphasis is on nonlinear dynamical systems approaches and dynamic (neural) fields.

Teaching Foundations of AIWolfram Burgard (Freiburg)

This course will introduce basic concepts and techniques used within the field of Artificial Intelligence. Among other topics, we will discuss:

  • Introduction to and history of Artificial Intelligence,
  • agents,
  • problem solving and search,
  • logic and knowledge representation,
  • planning
  • representation of and reasoning with uncertainty
  • machine learning
The Shark machine learning library (video lecture)Tobias Glasmachers, Christian Igel (Kopenhagen)

The Shark machine learning library is a modular C++ library for the design and optimization of adaptive systems. The library provides methods for regression, classification, and density estimation, including various kinds of neural networks and kernel methods, as well as general algorithms for nonlinear optimization, in particular single- and multi-objective evolutionary algorithms and gradient-based methods. Shark can be downloaded here.

Machine LearningStefan Kramer (Mainz)

Topics

  • decision trees: representation, learning, overfitting, pruning
  • ensembles: boosting, bagging, stacking, random forests
  • linear models: linear regression, ridge regression, logistic regression
  • neural networks: perceptron, multi-layer perceptron, back propagation
  • instance-based learning: k-NN, locally weighted learning, RBF networks, case-based reasoning
  • SVMs: margins, kernels
  • relational learning, inductive logic programming
  • reinforcement learning
  • genetic algorithms
Machine learningKatharina Morik (Dortmund)

Kaum ein Teilgebiet der Künstlichen Intelligenz hat sich so rasant entwickelt wie das maschinelle Lernen, dessen Methoden in vielen erfolgreichen Programmen (z.B. Google, Amazon) integriert sind und dessen Ergebnisse für viele Anwendungen (z.B. Marketing, Medizin) erfolgreich genutzt werden. Die Vorlesung behandelt die Lernaufgaben

  • Klassifikation
  • Subgruppenentdeckung
  • Merkmalsauswahl und -extraktion
  • Clustering.

Dabei werden verschiedene Methoden (Klassen von Algorithmen) mit ihrem jeweiligen theoretischen Hintergrund vorgestellt.

Machine Learning of Motor Skills for Robotics (video lecture)Jan Peters (Darmstadt)

Video lecture held at Deutsche Gesellschaft für Robotik (DGR-Tage 2011) in Karlsruhe.

Autonomous Intelligent Systems (video lecture)Cyrill Stachniss (Freiburg)

Video lecture held at Deutsche Gesellschaft für Robotik (DGR-Tage 2011) in Karlsruhe.

Machine Learning (video lecture)Andrew Ng (Stanford)

In this video lecture Professor Ng gives an overview to the course which provides a broad introduction to machine learning and statistical pattern recognition. Topics include supervised learning, unsupervised learning, learning theory, reinforcement learning and adaptive control. Recent applications of machine learning, such as to robotic control, data mining, autonomous navigation, bioinformatics, speech recognition, and text and web data processing are also discussed.

Complete Playlist for the Course:
http://www.youtube.com/view_play_list?p=A89DCFA6ADACE599

Custom Tables Plugin 05a1a29bdcae7b12229e651a9fd48b11