Competitive learning with generalized winner-take-all activation.

IEEE Trans Neural Netw

Dept. of Electr. and Comput. Eng., Carnegie Mellon Univ., Pittsburg, PA.

Published: October 2012

Competitive learning paradigms are usually defined with winner-take-all (WTA) activation rules. The paper develops a mathematical model for competitive learning paradigms using a generalization of the WTA activation rule (g-WTA). The model is a partial differential equation (PDE) relating the time rate of change in the ;density' of weight vectors to the divergence of a vector field called the neural flux. Characteristic trajectories are used to study solutions of the PDE model over scalar weight spaces. These solutions show how the model can be used to design competitive learning algorithms which estimate the modes of unknown probability density functions.

Download full-text PDF

Source
http://dx.doi.org/10.1109/72.125858DOI Listing

Publication Analysis

Top Keywords

competitive learning
16
learning paradigms
8
wta activation
8
competitive
4
learning generalized
4
generalized winner-take-all
4
winner-take-all activation
4
activation competitive
4
paradigms defined
4
defined winner-take-all
4

Similar Publications

To ensure their survival, animals must be able to respond adaptively to threats within their environment. However, the precise neural circuit mechanisms that underlie flexible defensive behaviors remain poorly understood. Using neuronal manipulations, machine learning-based behavioral detection, electron microscopy (EM) connectomics and calcium imaging in Drosophila larvae, we map second-order interneurons that are differentially involved in the competition between defensive actions in response to competing aversive cues.

View Article and Find Full Text PDF

Objective: To evaluate the psychological reactions, perceptions and opinions of cardiovascular preparticipation screening (PPS) among young competitive athletes.

Methods: This convergent parallel mixed-methods study recruited 222 athletes (mean age: 18.7 years) who underwent PPS at a Canadian university.

View Article and Find Full Text PDF

Motivation: Machine learning-based scoring functions (MLBSFs) have been found to exhibit inconsistent performance on different benchmarks and be prone to learning dataset bias. For the field to develop MLBSFs that learn a generalisable understanding of physics, a more rigorous understanding of how they perform is required.

Results: In this work, we compared the performance of a diverse set of popular MLBSFs (RFScore, SIGN, OnionNet-2, Pafnucy, and PointVS) to our proposed baseline models that can only learn dataset biases on a range of benchmarks.

View Article and Find Full Text PDF

In the highly competitive, perennially-changing healthcare industry, management prowess is vitally needed by those in positions of authority, but in isolation it is insufficient for delivering optimal results. Instead, management proficiencies must be complemented by leadership skills which amplify managerial efforts and outcomes. One avenue for building leadership acumen in healthcare organizations involves the establishment of leadership development programs.

View Article and Find Full Text PDF

Background/purpose: Our previous study found that the integration of arts into the dental education through an innovative photomicrograph competition can increase the dental students' interests in learning microscopic lessons. This study aimed to validate whether the same photomicrograph competition still had its effectiveness to promote the students' interests in learning microscopic lessons.

Materials And Methods: A photomicrograph competition was organized for the 477 dental students.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!