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

Raven's Progressive Matrices (RPM) is highly correlated with humanintelligence, and it has been widely used to measure the abstract reasoningability of humans. In this paper, to study the abstract reasoning capability ofdeep neural networks, we propose the first unsupervised learning method forsolving RPM problems. Since the ground truth labels are not allowed, we designa pseudo target based on the prior constraints of the RPM formulation toapproximate the ground truth label, which effectively converts the unsupervisedlearning strategy into a supervised one. However, the correct answer is wronglylabelled by the pseudo target, and thus the noisy contrast will lead toinaccurate model training. To alleviate this issue, we propose to improve themodel performance with negative answers. Moreover, we develop adecentralization method to adapt the feature representation to different RPMproblems. Extensive experiments on three datasets demonstrate that our methodeven outperforms some of the supervised approaches. Our code is available athttps://github.com/visiontao/ncd.

Unsupervised Abstract Reasoning for Raven's Problem Matrices