A Hybrid KNN–Random Forest System for Electronic Employee Effectiveness Evaluation with AHP-Based Criteria Formation
Keywords:
employee performance evaluation, KPI, hybrid classification, K-Nearest Neighbours, Random Forest, confidence routing, Analytic Hierarchy Process, HR analyticsAbstract
Organisations need employee evaluation tools that are objective, fast and adaptable to their own structure. Existing systems rely on fixed criteria, rigid database schemas, and algorithms that trade speed for accuracy. This study presents an electronic employee effectiveness evaluation system built on a hybrid classifier. K-Nearest Neighbours (KNN) classifies every employee and returns a confidence score; cases with confidence at or below a calibrated threshold θ = 0.72 are routed to a Random Forest (RF). Evaluation criteria are formed with Analytic Hierarchy Process (AHP) pairwise weighting and Pearson correlation screening, and stored in a ten-entity relational schema in Third Normal Form. Compared with the single-algorithm baseline, the hybrid model raised accuracy from 76.2% to 92.8%, F1-score from 74.7% to 90.9%, and cut processing time from 12.4 s to 6.1 s (−50.8%). At θ = 0.72, 78% of evaluations stay on the fast KNN path. The system also includes a KPI suggestion engine and a conversational AI advisor for HR specialists
References
1. DeNisi A. S., Murphy K. R. Performance appraisal and performance management: 100 years of progress? Journal of Applied Psychology, 2017, 102(3), 421–433.
2. Aguinis H. Performance Management. 4th ed. Chicago: Chicago Business Press, 2019.
3. Drucker P. F. The Practice of Management. New York: Harper & Row, 1954.
4. Kaplan R. S., Norton D. P. The balanced scorecard: measures that drive performance. Harvard Business Review, 1992, 70(1), 71–79.
5. Cover T., Hart P. Nearest neighbor pattern classification. IEEE Transactions on Information Theory, 1967, 13(1), 21–27.
6. Breiman L. Random forests. Machine Learning, 2001, 45(1), 5–32.
7. Saaty T. L. The Analytic Hierarchy Process. New York: McGraw-Hill, 1980.
8. Kohavi R. A study of cross-validation and bootstrap for accuracy estimation and model selection. Proc. IJCAI-95, 1995, 1137–1143.
9. Codd E. F. A relational model of data for large shared data banks. Communications of the ACM, 1970, 13(6), 377–387.
10. Pedregosa F. et al. Scikit-learn: machine learning in Python. Journal of Machine Learning Research, 2011, 12, 2825–2830.
11. Zokirov S. I., Karimova G. A. Mehnat samaradorligi mezonlarini aniqlash va baholash tizimi [Employee effectiveness criteria determination and evaluation system]. Certificate of official registration of a computer program No. DGU 53438. Ministry of Justice of the Republic of Uzbekistan, 2025.
12. Zokirov S. I., Karimova G. A. Development of a system for forming criteria for determining employee performance (KPI) and methods of calculating the assessment [in Russian]. Engineering Problems and Innovations, 2025, 3(25). ISSN 2181-4171

