Aiming at the problem of poor interpretability of tactical decision-making caused by the black box nature of machine learning,the intelligent decision-making technology of tactical decision prediction based on interpretable machine learning is studied.By expounding the concept of tactical prediction and combining with the typical characteristics of tactical prediction,a prediction model is constructed,and the advantages of using interpretable supervised learning method for intelligent decision-making are introduced.The principle of interpretable tree model and intelligent optimization algorithm are expounded,and the research status in the field of tactical prediction is analyzed.From the point of view of the combination of the two,the promotion of tactical prediction is verified,and the future development trend of interpretable machine learning is prospected.