Evaluation Metrics Research for Explainable Artificial Intelligence Global Methods Using Synthetic Data

In recent years, artificial intelligence technologies have been developing more and rapidly, a lot of research is aimed at solving the problem explainable intelligence. Various XAI methods are being developed to allow user understand logic how machine learning models work, in order compare methods, it necessary evaluate them. The paper analyzes various approaches evaluation defines requirements for system suggests metrics determine technical characteristics methods. A study was conducted, using these metrics, which determined degradation explanation quality SHAP LIME with increasing correlation input data. Recommendations also given further field practical implementation expanding scope their use.