Employee Performance Prediction: An Integrated Approach of Hybrid Improved LA-CNN-SVM-BiLSTM Model

Successful companies have a knack for predicting how their employees will perform in the future. Because people are the most valuable asset every company has and because their contributions decide the company's fate, decision-makers and managers across all industries are keenly interested in developing strategies to assemble a team of specialists. Important elements include feature selection, model training, and preprocessing. Normalization and discritization are the two preprocessing steps; discritization provides the best accuracy. The process of feature selection involves evaluating features based on their capacity to maintain data similarity. More specifically, the proposed approach used the Improved LA-CNN-SVM-BiGRU framework to train the model. In comparison to other methods, such as CNN and SVM, the suggested method obtains a better accuracy (about 95.72%).