Data mining refers to extracting knowledge by analyzing data from different perspectives. Data mining tools predict future trends and behaviors, allowing businesses to make proactive, knowledge driven decisions. The data is collected from different sources from different fields depending upon type of knowledge to be acquired. Classification is one of the basic data mining operations which aims to predict categorical data labels and helps in decision making. Ant-Miner is bio-inspired, ant colony based classification rules generator which is found much effective than other classifiers. Ant- Miner cannot be beneficial for a very large dataset. This paper can improve the Ant- Miner by using the concept of k-mean clustering, so that it can be applied on large data set. In this paper, data set “Weight Lifting Exercises monitored with Inertial Measurement Units Data Set” is used taken from “UCI repository”. The simulation results show the better performance of modified Ant- Miner as compared to existing Ant- Miner with reduced tree size.