Using Episodic Memories to Predict Upcoming Events

This paper addresses an important problem in control of episodic memory to be used to predict upcoming states in an environment where past situations sometimes reoccur. One of the key benefits is reducing the risk of retrieving irrelevant memories. Read more

To simulate the task of event processing, they define an event as a sequence of states, sampled from an underlying graph that represents the event schema. Figure below shows a ‘coffee shop visit’ event schema graph with three time points; each time point has two possible states. Each instance of an event (here, each visit to the coffee shop) is associated with a situation – a collection of features set to particular values; importantly, the features of the current situation deterministically control the transitions between states within the event.

This entry was posted in Decision Making, Decision Modeling, Event-driven, Knowledge Representation, Machine Learning. Bookmark the permalink.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s