A NOVEL AND EFFICIENT KNN USING MODIFIED APRIORI ALGORITHM

In the field of data mining, classification and association set rules are two of very important techniques to find out new patterns. K-nearest neighbor and apriori algorithm are most usable methods of classification and association set rules respectively. However, individually they face few challenges, such as, time utilization and inefficiency for very large databases. The current paper attempts to use both the methods hand in hand. Here, we have modified the apriori algorithm and used it to classify data for K-nearest neighbor. Modified Apriori helps in finding out only a few of the attributes that mainly define the class. These attributes are named as prominent attributes in this paper. This technique helps in improving the efficiency of KNN to a high extent.