Abstract
While substantial progress has been made to understand the functioning of the human brain from a computational perspective, little progress has been made in the area of memory storage and recall. This study develops a parameter-based computational model of long-term declarative episodic memory to extend Rosenblatt's proposed clock-system memory model by using deep learning networks with a novel one-shot learning algorithm. Experiments were conducted using the MNIST, CFAR100, Fashion MNIST, and A-Z datasets to show that the memory model can recall lifetime sequences of input images.