An Enhanced Classification Technique for Talent Management Using CACC-SVM

Classification of data is becoming a major challenge in Human Resource Management (HRM). The talent management problem in HRM is commonly solved through several classification techniques available in data mining. However the goal of the classification process is to classify the data in a highly accurate manner. Hence in this paper we propose a hybrid classification technique CACC-SVM for classifying data. The concept of discretization and classification are combined. This effectively increases the classification accuracy. The Class Attribute Contingency Coefficient (CACC) is a static, global, incremental, supervised & top down discretization algorithm. This produces concise summarization of continuous attributes which makes the classification process more accurate. The discretized data are then classified using high performing generalized classifier Support Vector Machine (SVM). The result of the proposed algorithm is compared with several traditional classification algorithms. Performance of the algorithms is measured through accuracy rate and error rate. The accuracy rates are higher and error rates are lower for the proposed algorithm.