LEM2 is basically an ready application for generating the rules of applied decision system.

It allows you to generate rules by "Learn from Examples by Modules", "Exhaustive" and "Sequential Covering" methods.

How it can be used? Lets say that decision system is an array of actions and decisions

Thanks to this knowledge (actions and decisions), an application is able to predict decisions taken by player according to his action

It has many purposes, for example you can use this algorithms for npc learning, police learning in race games, or in multiplayer games to predict your opponent moves

LEM2 asset is coded in C#. Package also contains a manual which guide you trhough the package.

For more information about how does the methods works, please refer to the following articles

LEM - machine learning

Sequential Covering

Exhaustive - Brute Force