User Tools

Site Tools


exploiting_cognitive_constraints_to_improve_machine-learning_memory_models

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

exploiting_cognitive_constraints_to_improve_machine-learning_memory_models [2015/12/17 21:59] (current)
Line 1: Line 1:
 +{{tag>"​Talk Summaries"​ "​Neural Networks"​ "NIPS 2015"​}}
  
 +====== Exploiting Cognitive Constraints to Improve Machine-Learning Memory Models ======
 +
 +| Presenter | Michael Mozer |
 +| Context | NIPS 2015 Reasoning, Attention, and Memory Workshop |
 +| Date | 12/12/15 |
 +
 +The human visual system has provided a good inspiration for computer visual system, so maybe the memory system can provide inspiration for reasoning attention and memory systems. ​ Understanding human memory is important for machine learning systems which must predict what information is interesting or available at a point in time.  For example, given a simple task where one stimulus requires a response, and a different stimulus requires another response, and stimuli are presented sequentially,​ the previous trials can have a strong effect on the response time.  This is a reflection of a simple memory which looks at the recent past when building up an expectation about the next stimulus. ​ In some experiments it has been suggested that the dependence on the past follows a power law, which is a property of explicit sorts of human memory (e.g. studying to learn facts which will be tested at a later point). ​ It has been shown that spaced exposure to information results in better retention over a longer amount of time.  The relationship between the spacing of exposure and the required retention period is non-uniform and non-monotonic among individual retention pyramids. ​ The spacing between exposure sessions and retention interval roughly follows a power law fit, and can be modeled e.g. as a neural network or a cascade of leaky integrators or a Kalman filter. ​ Some key features of these models are that when an event happens, a memory of it is stored in multiple traces (intervals),​ with the traces decaying at different rates, and the memory strength is a weighted sum of traces, with the slower scales weighted less importantly. ​ This allows it to predict the amount to which we might remember a stimulus based on how long ago it was, how often we were exposed to it, and how many times. ​ Compared to gated recurrent network models (LSTMs), the recurrent networks have little or no decay. ​ Compared to networks with learned decay constants, the networks do not have an enforced dominants of faster times scales. ​ Hierarchical recurrent networks have fixed decay constants, and history compression networks have their compression event based instead of time-based. ​ A model combining all of the necessary components to mimic human memory would be one where a number of groups of recurrent units with fixed decay rates, with a learned input mapping. ​ An appropriate model may be able to predict human memory and behavior better.
exploiting_cognitive_constraints_to_improve_machine-learning_memory_models.txt ยท Last modified: 2015/12/17 21:59 (external edit)