论文标题
用神经网络解决乌鸦的进步矩阵
Solving Raven's Progressive Matrices with Neural Networks
论文作者
论文摘要
Raven的进步矩阵（RPM）已被广泛用于人类的智能（IQ）测试。在本文中，我们旨在以受监督和无监督的方式与神经网络解决RPM。首先，我们研究了减少监督学习中过度合适的策略。我们建议在大规模数据集中使用具有深层和预训练的神经网络以改善模型概括。 Raven数据集的实验表明，我们监督方法的总体准确性超过了人类水平的性能。其次，由于智能代理需要自动学习解决新问题的新技能，因此我们提出了第一个无监督方法，即具有伪目标（MCPT）的多标签分类，以解决RPM问题。根据伪目标的设计，MCPT将无监督的学习问题转换为监督任务。实验表明，MCPT将随机猜测的测试精度增加一倍，例如28.50％比12.5％。最后，我们讨论了将来用无监督和可解释的策略解决RPM的问题。
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.
