Publication at Faculty of Mathematics and Physics |
2009
Abstract
Associative memories represent a model of artificial neural networks applicable to the information storage and retrieval. However, the performance of traditional associative memories is very sensitive to the number of stored patterns and their mutual similarities.
In order to avoid limitations imposed by processing larger amounts of mutually correlated patterns, we have developed the so-called Hierarchical Associative Memory (HAM) model. This paper is focused on the time complexity and memory complexity of the HAM model.
The time complexity of the HAM model is derived. The memory complexity is analyzed and the theoretical results are compared with the experimental results.