Data classification with dynamically growing and shrinking neural networks

doi:10.1016/j.jocs.2025.102660

Data classification with dynamically growing and shrinking neural networks

2025-07-01

0

PRE

AI

S

Szymon Świderski

A

Agnieszka Jastrzębska *

DOI：10.1016/j.jocs.2025.102660

Original

Original request for help

Save

Abstract

En 中文

• We show a procedure allowing dynamic shrinking and growing of a neural architecture. • Monte Carlo tree search simulates network behavior and predicts its performance. • The growing and shrinking procedure adjusts the model during training. • Proposed model was delivered for convolutional and feed-forward neural networks. • We apply the proposed model to multivariate time series classification.

Data classification with dynamically growing and shrinking neural networks

Abstract

Journal

Organization