A new method for employee promotion detection based on the hybrid ANN-SVM model

A company’s ability to achieve its ultimate long-term goal highly depends on the potential and motives of its employees. The Human Resource Management team must be able to accurately evaluate a worker’s ability in quality work and incentivize him/her using promotion. While it has been quite difficult to measure an employee’s aptness for promotion, previously, machine learning classification methods has been researched upon. The aim of this study is to analyze a newly proposed hybrid classification model in detecting potential employee for promotion. The hybrid model combines the strength in feature selection of Artificial Neural Network model and the binary classification power of Support Vector Machine. The model is tested and trained after its hyperparameters are tuned. Afterward, the model is evaluated on the accuracy, precision, recall and F1 score, and compared to the individual SVM and ANN models for performance. The predictions of this proposed model have proven to exceed those two models with a range of 0.15% to 7.26% increase in performances. Note, the precision of ANN model is somewhat higher than the hybrid model. In summary, the hybrid ANN-SVM model has proven to work effectively with both imbalanced and balanced datasets.