This paper proposes a hybrid genetic algorithm (HGA) for feature selection, which integrates local search operations into a simple genetic algorithm (GA) to enhance its performance. The hybridization technique improves the final performance and enables subset-size control. The hybrid GAs showed better convergence properties compared to classical GAs. A rigorous timing analysis was developed to compare the timing requirements of conventional and proposed algorithms. Experiments on various standard datasets showed that the proposed hybrid GA outperformed both a simple GA and sequential search algorithms in terms of accuracy, especially for large-sized problems. The feature selection problem involves selecting a subset of features from a larger set based on an optimization criterion. The goal is to design a more compact classifier with minimal performance degradation. The optimal solution is computationally intractable due to the exponential search space, so most algorithms lead to suboptimal solutions. The SFFS algorithm is considered the best among sequential search algorithms, but no clear-cut comparison exists between SFFS and GA. The hybrid GA improves the performance of the simple GA by embedding local search operations. These operations move solutions toward local optima and accumulate improvements over generations, leading to significant performance gains. The local search operations are parameterized and analyzed for their fine-tuning power and timing requirements. The hybrid GA outperforms other algorithms in terms of accuracy and computational efficiency. The paper discusses various feature selection algorithms, including enumeration algorithms, sequential search algorithms, and genetic algorithms. It reviews comparative studies and highlights the limitations of simple GAs, such as inferior solutions compared to classical heuristic algorithms. The hybrid GA addresses these limitations by incorporating domain-specific knowledge through local search operations. The hybrid GA is designed to solve the feature selection problem by embedding local search operations into the simple GA. The local search operations are defined and analyzed for their effectiveness and timing requirements. The hybrid GA is tested on various datasets and shows superior performance in terms of classification accuracy and computational time. The paper also discusses the timing analysis of local search operations, showing that the ripple factor influences the number of atomic operations and the speedup factor. The hybrid GA is compared with other algorithms, including the multistart algorithm, and shows better performance in terms of convergence and solution quality. The results indicate that the hybrid GA is more effective than the simple GA and SFFS, especially for large-sized problems. The paper concludes that the hybrid GA provides better overall performance than the simple GA and SFFS, particularly for large-sized problems.This paper proposes a hybrid genetic algorithm (HGA) for feature selection, which integrates local search operations into a simple genetic algorithm (GA) to enhance its performance. The hybridization technique improves the final performance and enables subset-size control. The hybrid GAs showed better convergence properties compared to classical GAs. A rigorous timing analysis was developed to compare the timing requirements of conventional and proposed algorithms. Experiments on various standard datasets showed that the proposed hybrid GA outperformed both a simple GA and sequential search algorithms in terms of accuracy, especially for large-sized problems. The feature selection problem involves selecting a subset of features from a larger set based on an optimization criterion. The goal is to design a more compact classifier with minimal performance degradation. The optimal solution is computationally intractable due to the exponential search space, so most algorithms lead to suboptimal solutions. The SFFS algorithm is considered the best among sequential search algorithms, but no clear-cut comparison exists between SFFS and GA. The hybrid GA improves the performance of the simple GA by embedding local search operations. These operations move solutions toward local optima and accumulate improvements over generations, leading to significant performance gains. The local search operations are parameterized and analyzed for their fine-tuning power and timing requirements. The hybrid GA outperforms other algorithms in terms of accuracy and computational efficiency. The paper discusses various feature selection algorithms, including enumeration algorithms, sequential search algorithms, and genetic algorithms. It reviews comparative studies and highlights the limitations of simple GAs, such as inferior solutions compared to classical heuristic algorithms. The hybrid GA addresses these limitations by incorporating domain-specific knowledge through local search operations. The hybrid GA is designed to solve the feature selection problem by embedding local search operations into the simple GA. The local search operations are defined and analyzed for their effectiveness and timing requirements. The hybrid GA is tested on various datasets and shows superior performance in terms of classification accuracy and computational time. The paper also discusses the timing analysis of local search operations, showing that the ripple factor influences the number of atomic operations and the speedup factor. The hybrid GA is compared with other algorithms, including the multistart algorithm, and shows better performance in terms of convergence and solution quality. The results indicate that the hybrid GA is more effective than the simple GA and SFFS, especially for large-sized problems. The paper concludes that the hybrid GA provides better overall performance than the simple GA and SFFS, particularly for large-sized problems.