سينا خرمي

A major challenge in machine learning algorithms and learning methods is their lack of conformity with learning in humans. Learning in humans is done by theories of cognitive psychologists. In all of these cognitive theories of memory plays a very important role in learning and calling concepts. Associative memories are very useful in these theories due to the establishment of relationships between the various concepts. For example, we see the names of the acquaintances who saw the face of the acquaintances, where we made an affinity between the face and the name. In this project, inspired by cognitive learning theories and associative memories, a new method of learning the machine based on the method of learning the human mind was designed and implemented that features such as the ability to learn incremental and the ability to learn a variety of concepts with different dimensions. an‎d the relationships between them, as well as a lot of learning abilities. The obtained algorithm was compared quantitatively using conventional data of binary text characters with other available methods and by qualitative evaluation of other experiments. The results of the evaluations performed in comparison with other algorithms in this area have shown remarkable progress in calling noise data and disturbances.