Comparative Evaluation of Data Mining Classification Techniques for Assistant Recruitment in a Computer Laboratory

This study analyzed different data mining classification techniques for predicting assistant acceptance decisions in a computer lab's recruitment process.Historical data was utilized on 128 accepted and 150 rejected candidates over 4 years.After pre-processing, classifiers like decision trees, Naïve Bayes, and K-Nearest Neighbor were implemented using Python for comparative evaluation.Accuracy metrics revealed KNN as optimal with 92.86% correct prediction over alternatives.However, precision and false positive rates require additional scrutiny from a recruitment perspective.The findings imply that customized algorithm tuning and balanced error analysis allow better leveraging of predictive recruiting to minimize selection inaccuracies despite growing applicant volumes.Though specific to an organizational context, results initiate valuable trajectories for effective data-driven human resources practice.