The paper introduces a modified Crayfish Optimization Algorithm (MCOA) to address the limitations of the original Crayfish Optimization Algorithm (COA), which includes decreased search efficiency and a tendency to fall into local optima. MCOA incorporates an environmental renewal mechanism inspired by crayfish survival habits, using water quality factors to guide crayfish towards better environments. Additionally, it integrates a learning strategy based on ghost antagonism to enhance its ability to avoid local optima. The performance of MCOA was evaluated using the IEEE CEC2020 benchmark functions and four constrained engineering problems and feature selection problems. For the constrained engineering problems, MCOA improved results by 11.16%, 1.46%, 0.08%, and 0.24% compared to COA. For feature selection problems, the average fitness value and accuracy improved by 55.23% and 10.85%, respectively. The combination of the environment updating mechanism and the learning strategy based on ghost antagonism significantly enhances MCOA's performance, making it suitable for solving complex spatial and practical application problems. This research has important implications for the field of optimization.The paper introduces a modified Crayfish Optimization Algorithm (MCOA) to address the limitations of the original Crayfish Optimization Algorithm (COA), which includes decreased search efficiency and a tendency to fall into local optima. MCOA incorporates an environmental renewal mechanism inspired by crayfish survival habits, using water quality factors to guide crayfish towards better environments. Additionally, it integrates a learning strategy based on ghost antagonism to enhance its ability to avoid local optima. The performance of MCOA was evaluated using the IEEE CEC2020 benchmark functions and four constrained engineering problems and feature selection problems. For the constrained engineering problems, MCOA improved results by 11.16%, 1.46%, 0.08%, and 0.24% compared to COA. For feature selection problems, the average fitness value and accuracy improved by 55.23% and 10.85%, respectively. The combination of the environment updating mechanism and the learning strategy based on ghost antagonism significantly enhances MCOA's performance, making it suitable for solving complex spatial and practical application problems. This research has important implications for the field of optimization.