人工智能机理解释与数学方法探讨

Artificial intelligence mechanism and mathematics implementation methods

人工智能是一门交叉学科,融合了脑认知科学、心理学、数学(包括统计学)、信息科学、计算机科学等诸多学科.目前,以大数据驱动的深度学习为代表的人工智能发展迅速,在许多领域有广泛的应用.但是,人工智能现在最大的缺陷是没有理论,基本上还处于实验科学阶段.探讨人工智能的机理,进行其理论研究,对人工智能的进一步发展至关重要.为了揭示人工智能的本质,进一步促进人工智能的理论研究,本文从认知生物学和认知心理学两个层面,探讨现有的三个主要的人工智能学派—联结主义、符号主义和行为主义—所蕴含的机理和具体实现的数学理论和方法,试图解释人工智能的本质,给出未来可解释的和具有创造性的人工智能的一些可能的基础研究方法和研究方向.

Artificial intelligence is an interdisciplinary discipline that integrates many subjects, such as brain cognitive science, psychology, mathematics(including statistics), information science, and computer science. At present, artificial intelligence represented by big data-driven deep learning has developed rapidly and has been widely used in many fields. However, the biggest flaw in artificial intelligence right now is that it has no theory and is basically still in the experimental science stage. To explore the mechanism of artificial intelligence and carry out its theoretical research is of vital importance to the further development of artificial intelligence. To reveal the nature and promote the theoretical research of artificial intelligence, the mechanisms and concrete realizations of the three existing artificial intelligence schools are explored from the two levels of cognitive biology and cognitive psychology, such as symbolism, connectionism and behaviorism, trying to explain the nature of artificial intelligence. Some possible basic research methods and research directions of future interpretable and creative artificial intelligence are proposed.