Solving Raven's Progressive Matrices with Neural Networks

Abstract

Raven's Progressive Matrices (RPM) have been widely used for Intelligence Quotient (IQ) test of humans. In this paper, we aim to solve RPM with neural networks in both supervised and unsupervised manners. First, we investigate strategies to reduce over-fitting in supervised learning. We suggest the use of a neural network with deep layers and pre-training on large-scale datasets to improve model generalization. Experiments on the RAVEN dataset show that the overall accuracy of our supervised approach surpasses human-level performance. Second, as an intelligent agent requires to automatically learn new skills to solve new problems, we propose the first unsupervised method, Multilabel Classification with Pseudo Target (MCPT), for RPM problems. Based on the design of the pseudo target, MCPT converts the unsupervised learning problem to a supervised task. Experiments show that MCPT doubles the testing accuracy of random guessing e.g. 28.50% vs. 12.5%. Finally, we discuss the problem of solving RPM with unsupervised and explainable strategies in the future.

References in corpus (5)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting
• Measuring abstract reasoning in neural networks
• Learning Perceptual Inference by Contrasting
• IQ of Neural Networks

Cited by in corpus (8)

• Abstraction and Analogy-Making in Artificial Intelligence
• Deep Learning Methods for Abstract Visual Reasoning: A Survey on Raven's Progressive Matrices
• Effective Abstract Reasoning with Dual-Contrast Network
• Pointer Value Retrieval: A new benchmark for understanding the limits of neural network generalization
• Multi-Granularity Modularized Network for Abstract Visual Reasoning
• Pairwise Relations Discriminator for Unsupervised Raven's Progressive Matrices
• Analogical and Relational Reasoning with Spiking Neural Networks
• Unsupervised Abstract Reasoning for Raven's Problem Matrices