Forecasting Employee Potential through Probationary Assessment

Abstract

Effective corporate governance necessitates the continual nurturing and cultivation of employee potential for long-term professional success. However, assessing an employee's potential and performance objectively and consistently from the start of their career presents a substantial difficulty in reducing any mismatches with the company's goals and expectations. This study introduces a predictive methodology that uses probationary employee performance to map their potential. The study focuses on Performance (Y-axis) and Potential (X-axis) variables using data from 265 employees at Company X who went through a probationary period. Various machine learning models, including Logistic Regression, Naive Bayes, k-NN, SVM, and Decision Tree, were used to analyze data using Orange Data Mining software. The Logistic Regression model has the highest accuracy, at 90% (0.906). Validity testing, using the Confusion Matrix, allowed individuals to be classified into nine potential groups, in accordance with the 9-Box Matrix Talent Management paradigm. This classification provides HR with a strategic tool for tailoring career development strategies based on expected potential within their respective sectors.