Statistical physics of learning vector quantization
PhD ceremony: Mr. A.W. Witoelar, 11.00 uur, Academiegebouw, Broerstraat 5, Groningen
Thesis: Statistical physics of learning vector quantization
Promotor(s): prof. M. Biehl, prof. N. Petkov
Faculty: Mathematics and Natural Sciences
The field of machine learning concerns the design of algorithms to learn and recognize complex patterns from data. Learning Vector Quantization (LVQ) constitutes an important family of such algorithms using prototypes which serve as typical examples. Despite its wide range of applications, the theoretical understanding of LVQ in general remains very limited. In this thesis a theoretical framework is constructed using concepts from statistical physics which allows for an exact analysis of the typical learning behaviors of the system. Studies in this thesis compare the characteristics of various LVQ algorithms to demonstrate the robustness of Neural Gas (NG) schemes and the specific advantages of LVQ 2.1 and Robust Soft LVQ algorithms. Furthermore surprising non-trivial behaviors are revealed including learning plateaus and phase transitions in the training process. The results provide insights to general prototype-based learning prescriptions.
Last modified: | 13 March 2020 01.13 a.m. |
More news
-
07 February 2025
Four FSE start-ups take part in Academic Startup Competition
Four start-up companies founded by FSE scientists have been nominated to participate in the fifth edition of the Academic Start-up Competition.
-
03 February 2025
Flexible solar panels and a tattooed sensor
The black dots and lines on this plastic plate conduct electricity, and are as flexible as the plastic it is on. And that is special, Ranjita Bose, associate professor of Polymer Engineering, explains: ‘It’s a conductive polymer that combines the...
-
30 January 2025
Highlighted papers December 2024 - January 2025
Read our highlighted papers from December/January: new insights into electronics from 2D materials, and into the protein clumps that cause Huntington's disease.