Patients undergoing a bone marrow stem cell transplant face various risk factors, stem cell is a procedure to replace the damaged or destroyed bone marrow with healthy bone marrow. Stem cells are immature cells in the bone marrow that gives rise to all of blood cells. Totally 2000 patients records are collected also split into training and test data of 768 records with 18 different attributes. Non-matrix factorization algorithm are used to find the missing values, for handling the more attributes and more data set of different patients with excluding of errors and missing values. The data set focused in targeted information extraction and investigative analysis along with useful patterns.  Various classification algorithms like SVM, RF, and NN are trained on predicting the survival of each patient depends on their preoperative measurements along with highest prominence. Non-matrix algorithm increases the accuracy of prediction result.