Topological Data Analysis in Conjunction with Traditional Machine Learning Techniques to Predict MDAP PM Ratings

Abstract

Topological data analysis (TDA) is an unconventional machine learning technique that is used to understand the underlying topology of data. The premise is that data has shape. The two methodologies used in TDA are persistent homology and the mapper algorithm. Traditional machine learning techniques include supervised unsupervised methods such as clustering, Bayesian networks, neural networks, support vector machines (SVM), and random forests. The goal of this study is to apply TDA methods in conjunction with traditional machine learning algorithms to Defense Acquisition Executive Summary (DAES) data to determine if TDA helps to improve prediction measures (accuracy, f-measure, sensitivity, and specificity) over using traditional methods only when predicting program manager ratings from Major Defense Acquisition Programs (MDAPs). We show that TDA when used in conjunction with traditional machine learning models at a local level of the DAES data improved the accuracy of predicting PM cost ratings of MDAPs at 80% of all nodes in training and testing as compared to implementing these models without TDA at the global level.