At present, there are many shortcomings of genetic optimization algorithm, including the complexity of genetic algorithm programming, involving gene coding and decoding, the setting of crossover rate, the mutation rate and other parameters in the algorithm determined by experience and the intense dependence on the merits of the initial population. In the financial performance evaluation, genetic optimization algorithm needs to artificially add and modify the evaluation indicators, and the new indicators can not increase the evaluation according to the situation, resulting in a large gap between the evaluation indicators. The data produced in the financial performance evaluation of enterprises is not accurate enough to achieve the optimal financial performance. On the basis of genetic algorithm, simulated anneal algorithm and machine learning technology were made up of a hybrid genetic optimization algorithm, which improved the population diversity of genetic algorithm, the intelligent data learning ability of genetic optimization algorithm and the efficiency of algorithm data decomposition and processing. After data comparison and analysis, it was finally determined that the improved hybrid genetic optimization algorithm was more accurate and intelligent in data analysis than the traditional genetic optimization algorithm, and the financial performance was also the best. For example, the highest rate of return of the improved algorithm in investment projects was 80%, and the highest accuracy of data analysis in the optimized financial performance system was 95%, which showed the effectiveness of the hybrid genetic optimization algorithm. The improved hybrid genetic optimization algorithm could indeed make the financial performance evaluation indicators more intelligent and efficient in the enterprise financial performance evaluation system and achieve better results in the field of financial performance evaluation.
