English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Feature selection aims to select a subset of features to optimize theperformances of downstream predictive tasks. Recently, multi-agent reinforcedfeature selection (MARFS) has been introduced to automate feature selection, bycreating agents for each feature to select or deselect corresponding features.Although MARFS enjoys the automation of the selection process, MARFS suffersfrom not just the data complexity in terms of contents and dimensionality, butalso the exponentially-increasing computational costs with regard to the numberof agents. The raised concern leads to a new research question: Can we simplifythe selection process of agents under reinforcement learning context so as toimprove the efficiency and costs of feature selection? To address the question,we develop a single-agent reinforced feature selection approach integrated withrestructured choice strategy. Specifically, the restructured choice strategyincludes: 1) we exploit only one single agent to handle the selection task ofmultiple features, instead of using multiple agents. 2) we develop a scanningmethod to empower the single agent to make multiple selection/deselectiondecisions in each round of scanning. 3) we exploit the relevance to predictivelabels of features to prioritize the scanning orders of the agent for multiplefeatures. 4) we propose a convolutional auto-encoder algorithm, integrated withthe encoded index information of features, to improve state representation. 5)we design a reward scheme that take into account both prediction accuracy andfeature redundancy to facilitate the exploration process. Finally, we presentextensive experimental results to demonstrate the efficiency and effectivenessof the proposed method.

Simplifying Reinforced Feature Selection via Restructured Choice Strategy of Single Agent