Explainable Machine Learning Framework for Image Classification Problems: Case Study on Glioma Cancer Prediction.

However, easily defined features such as size of object (e.g., tumor size in our case study), mean value of image pixels (e.g., tumor color tends to be black or white) and cyclic level (e.g., tumor shape tends to be cyclic or oval) are by default easily understandable to most humans but they may not be significant for the prediction problem which a machine learning model aims to solve.

also in 2,774 other papers