Towards Enhanced Decision Support through Learning from Past Experiences
Keyword(s): decision support, data mining, optimization, statistical learning
Abstract: The number of changes that IT departments have to deal with is growing at a fast pace in response to changing business needs of enterprises. As changes are getting executed and deployed, knowledge is being created and stored. It is of paramount importance to the success of the business to re-use that knowledge for future changes. In fact, those who do not learn from past experiences are doomed to repeat the same mistakes as well as not bear the fruit of the ones that were successful. This paper addresses this concern by providing for every change being worked out the most similar past changes. Our approach uses an optimization paradigm to model the problem of finding past similar changes by designing and learning similarity functions. Our approach enhances the efficiency and effectiveness of dealing with changes, by reducing the risk and shortening the time of introducing new changes.
External Posting Date: June 21, 2009 [Fulltext]. Approved for External Publication
Internal Posting Date: June 21, 2009 [Fulltext]