Solving Raven's Progressive Matrices with Multi-Layer Relation Networks
Abstract
Raven's Progressive Matrices are a benchmark originally designed to test the cognitive abilities of humans. It has recently been adapted to test relational reasoning in machine learning systems. For this purpose the so-called Procedurally Generated Matrices dataset was set up, which is so far one of the most difficult relational reasoning benchmarks. Here we show that deep neural networks are capable of solving this benchmark, reaching an accuracy of 98.0 percent over the previous state-of-the-art of 62.6 percent by combining Wild Relation Networks with Multi-Layer Relation Networks and introducing Magnitude Encoding, an encoding scheme designed for late fusion architectures.
Keywords
Cite
@article{arxiv.2003.11608,
title = {Solving Raven's Progressive Matrices with Multi-Layer Relation Networks},
author = {Marius Jahrens and Thomas Martinetz},
journal= {arXiv preprint arXiv:2003.11608},
year = {2020}
}
Comments
6 pages, 6 figures, to be published in the Proceedings of the IJCNN 2020, source code available at http://webmail.inb.uni-luebeck.de/exchange-supplement/PGM_MLRN_supplementary.zip