Abstract: A query optimizer is a core component of any Database Management System. Multiple approaches have been suggested which were based on framework of classical query evaluation procedures that heavily dependent on metadata. There are computational environments where metadata acquisition and support is very expensive. In this paper an optimization technique for complex queries using case based reasoning (machine learning technique) in ubiquitous computing environment is deliberated. In this technique a new problem is solved by finding a similar past case, and reusing it in the new problem situation. As CBR is an approach to incremental, sustained learning, since a new experience is retained each time a problem has been solved, making it immediately available for future problems which in return may create a bulky case base. Thus we were proposing a technique of dynamic deletion of irrelevant cases from case base. Through which system can detect the inappropriate cases and replace them with new case in order to maintain size of case base.