Institutional Repository of Vilnius University
Dirbtinių neuroninių tinklų taikymai besikeičiančio uždavinio sprendimui /
Abstract
dc:descriptionOne of the key challenges in the studies and application of artificial neural networks is the ability to adapt to a changing environment. Neural networks that have accumulated enough knowledge are the most efficient with the exact model they are trained on. However, if the model changes, the network will have to re-learn again. Neural networks are used in many different fields, like science, medicine, business or manufacturing. The changes in the model of learned network might might the need for learning from zero knowledge, which can lead to an expesive downtime of the network. This paper analyzes the described problem. Using the multilayer perceptron as an example, the best criteria for a better adaptation to a changing task is examined. The research includes the analy- sis of the architecture of multilayer perceptron, as well as its parameters. The gathered knowledge of optimal parameters is used to effectively solve a changing task.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2016
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Popok, Oleg,
- Contributors dc:contributor
-
- Valaitis, Vytautas
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD23223046&prefLang=en_US