Artificial Neural Networks

Artificial neural networks (ANN) were inspired by the architecture and function of the brain. Nevertheless, their greatest strength is not that they are good models of the brain, but rather that they are powerful function approximators. Since the 1980's many types of ANN have been developed and tricks for training ANNs on data proliferated. Recent advances in computing hardware and the availability of large datasets have made it possible to train ANNs such that they perform better than humans, e.g. on image recognition. In this class, students will, firstly, gain a theoretical understanding of the principles underlying the methods applied to neural networks and, secondly, learn practical skills in implementing neural networks and applying them for data analysis.

Topics: optimization problems, regression, logistic regression, biological neural networks, model selection, universal approximation theorem, perceptron, MLP, backpropagation, deep neural networks, recurrent neural networks, LSTM, Hopfield network, Bolzmann machine

Software: python, numpy, scipy, matplotlib, scikit-learn, tensorflow

There will be a written examination at the end of the course.

Lecturers

Details

Course type
Lectures
Credits
6 CP
Term
Winter Term 2020/2021
E-Learning
moodle course available

Dates

Lecture
Takes place every week on Monday from 16:00 to 18:00.
First appointment is on 26.10.2020
Last appointment is on 08.02.2021
Exercise
Takes place every week on Friday from 10:00 to 12:00.
First appointment is on 06.11.2020
Last appointment is on 12.02.2021
Tutorial
Takes place every week on Tuesday from 12:00 to 14:00.
First appointment is on 27.10.2020
Last appointment is on 09.02.2021
Tutorial
Takes place every week on Wednesday from 10:00 to 12:00.
First appointment is on 28.10.2020
Last appointment is on 10.02.2021

Requirements

Calculus, linear algebra, statistics, programming.

Documents

Link Moodle Course Page

The Institut für Neuroinformatik (INI) is a central research unit of the Ruhr-Universität Bochum. We aim to understand the fundamental principles through which organisms generate behavior and cognition while linked to their environments through sensory systems and while acting in those environments through effector systems. Inspired by our insights into such natural cognitive systems, we seek new solutions to problems of information processing in artificial cognitive systems. We draw from a variety of disciplines that include experimental approaches from psychology and neurophysiology as well as theoretical approaches from physics, mathematics, electrical engineering and applied computer science, in particular machine learning, artificial intelligence, and computer vision.

Universitätsstr. 150, Building NB, Room 3/32
D-44801 Bochum, Germany

Tel: (+49) 234 32-28967
Fax: (+49) 234 32-14210