The exponential rise of digital data presents significant data acquisition and evaluation challenges. Efficient techniques are needed to handle the large and complex datasets in today’s interconnected society. Traditional methods often struggle to derive meaningful insights from these vast data sets. We propose an Improved Genetic Algorithm (I-GA) integrated with a Support Vector Machine (SVM) and k-nearest Neighbors (kNN). This combination aims to efficiently extract features and accurately classify data from the CICIDS2017 dataset. The I-GA optimizes feature selection, while SVM and kNN handle the high-dimensional data classification. The I-GA-kNN model achieved superior performance with an accuracy of 97.71% and an F1 score of 97.53%. It outper- formed the I-GA-SVM model, which had an accuracy of 88.98% and an F1 score of 88.49%. The I-GA optimization process effectively reduces computational load, making it suitable for real-time applications. The hybrid I-GA approach significantly improves classification accuracy and efficiency. This makes it ideal for real-time applications by reducing computational overhead and enhancing an alytical dependability. The results underscore the I-GA algorithm’s capacity to augment the efficacy of diverse classifiers in classification endeavours. Our study demonstrates that the I-GA, combined with SVM and kNN, enhances classifier performance. This method offers a robust solution for extensive data analysis, with potential for further development and application. Future research should explore integrating deep learning techniques to improve accuracy and applicability across different fields.