Knowledge Discovery for Event Series Decision Based on Rough Set

To make decisions about event series is part of our life, and to discover knowledge from these decisions is of great significance in the field of controlling and decision-making. The paper takes event series as the exterior form of movements with the dynamic attributes, and gets the Markov transition probabilities matrix to express those attributes with statistics. First, according to the matrix, the decision table is constructed. Then, by reducing attributes based on rough set theory, the decision table is reduced, and the decision rules are acquired as well. Finally we make the decision through defining rule distance and taking the minimum rule distance as decision principle. Followed is an example, which proves that the algorithm is feasible and effective to the event series decision.