htw saar Piktogramm QR-encoded URL
Back to Main Page Choose Module Version:
XML-Code

flag

Introduction to the Mathematics of Neural Networks

Module name (EN):
Name of module in study programme. It should be precise and clear.
Introduction to the Mathematics of Neural Networks
Degree programme:
Study Programme with validity of corresponding study regulations containing this module.
Electrical Engineering, Master, ASPO 01.10.2013
Module code: E1924
SAP-Submodule-No.:
The exam administration creates a SAP-Submodule-No for every exam type in every module. The SAP-Submodule-No is equal for the same module in different study programs.
P211-0001
Hours per semester week / Teaching method:
The count of hours per week is a combination of lecture (V for German Vorlesung), exercise (U for Übung), practice (P) oder project (PA). For example a course of the form 2V+2U has 2 hours of lecture and 2 hours of exercise per week.
2V (2 hours per week)
ECTS credits:
European Credit Transfer System. Points for successful completion of a course. Each ECTS point represents a workload of 30 hours.
2
Semester: according to optional course list
Mandatory course: no
Language of instruction:
German
Assessment:
Project work

[updated 05.06.2025]
Applicability / Curricular relevance:
All study programs (with year of the version of study regulations) containing the course.

E1924 (P211-0001) Electrical Engineering and Information Technology, Master, ASPO 01.04.2019 , optional course, course inactive since 31.03.2020
E813 (P211-0001) Electrical Engineering, Master, ASPO 01.10.2005 , optional course
E1924 (P211-0001) Electrical Engineering, Master, ASPO 01.10.2013 , optional course
Workload:
Workload of student for successfully completing the course. Each ECTS credit represents 30 working hours. These are the combined effort of face-to-face time, post-processing the subject of the lecture, exercises and preparation for the exam.

The total workload is distributed on the semester (01.04.-30.09. during the summer term, 01.10.-31.03. during the winter term).
30 class hours (= 22.5 clock hours) over a 15-week period.
The total student study time is 60 hours (equivalent to 2 ECTS credits).
There are therefore 37.5 hours available for class preparation and follow-up work and exam preparation.
Recommended prerequisites (modules):
None.
Recommended as prerequisite for:
Module coordinator:
Prof. Dr. Harald Wern
Lecturer: Prof. Dr. Harald Wern

[updated 03.12.2012]
Learning outcomes:
Whenever the relationship between the input side and the output side cannot be explicitly specified by a function, neural networks are an interesting alternative.
After successfully completing this module, students will be able to configure a neural network and train a certain number of associations (epsilon) accuartely with at least a quadratic learning method.

[updated 05.06.2025]
Module content:
1. Basic information about neural networks, motivation and principles
2. Two-layer neural feed-forward networks
3. Three-layer neural feed-forward networks
4. Higher-order neural feed forward networks
5. Feedback networks for optimization tasks

[updated 05.06.2025]
Teaching methods/Media:
Overhead transparencies, video projector

[updated 05.06.2025]
Recommended or required reading:
Lenze, Burkhard: Einführung in die Mathematik neuronaler Netze
Scherer, Andreas: Neuronale Netze, Grundlagen und Anwendungen
Braun, Heinrich: Neuronale Netze, Optimierung durch Lernen und Evolution
Medsker, L.R.; Jain, L.C.: Recurrent Neural Networks, Design and Applications
Cichocki, A.; Unbehauen,R.: Neural Networks for Optimization and Signal Processing

[updated 05.06.2025]
 
[Fri Jul  4 08:00:11 CEST 2025, CKEY=eeidmnn, BKEY=em2, CID=E1924, LANGUAGE=en, DATE=04.07.2025]