Temporal Coding in Human Memory

Professor Kahana presented his newly developed Temporal Coding Model of human memory. This model is designed to explain behavioral data on human memory for time of occurrence as well as long term recency effects. The basic idea is that summing the vectors representing prior memories can be mathematically modeled as a random walk in a high dimensional space. This random walk has mathematical properties which enable it to represent temporal information. This information is then concatenated with the vectors representing newly experienced memories and stored in an auto associative neural network. Professor Kahana showed how this model overcomes a number of limitations in existing models of human memory.

Speaker Schedule  |  Reports from Previous Years
Top of Page | Life Sciences | Brandeis University