This study investigates the effect of rice husk ash (RHA) on compressive strength of pervious concrete and explores the use of machine learning (ML) for forecasting its strength. An inclusive dataset encompassing various parameters of pervious concrete with RHA was compiled from published research. This data was utilized to develop and assess ML models for predicting compressive strength. Six different algorithms, including Artificial Neural Network (ANN), Boosted tree regression (BT), K-nearest neighbors (KNN), Linear regression (LR), Support Vector Regression (SVR), and Extreme Gradient Boosting (XGB), were investigated. The findings indicate an optimal RHA content for achieving maximum strength, with compressive strength generally increasing to a 10% replacement level and then decreasing with further RHA substitution. The analysis showed that the SVR model was the most effective and reliable option for prediction. SVR model achieved greater performance related to other models, exhibiting a higher coefficient of determination and lower values for Root Mean Square Error and Mean Absolute Error. The study shows that SVR model can accurately identify how different factors in data influence each other. This makes it a valuable tool for predicting how strong pervious concrete is with RHA under compression. SHAP (SHapley Additive exPlanations) analysis showed that aggregate size significantly affects compressive strength, followed by water-to-binder ratio and curing period.