Comparison of machine learning techniques to predict all-cause mortality using fitness data: The Henry Ford exercIse testing (FIT) project

Medicine & Life Sciences

• Machine Learning 100%
• Exercise 57%
• Cardiorespiratory Fitness 57%
• Mortality 44%
• Decision Trees 36%
• Support Vector Machine 36%
• Forests 17%
• Health 14%
• Coronary Artery Disease 12%
• Medical Records 12%
• Area Under Curve 12%
• Datasets 11%
• Heart Failure 10%

Social Sciences

• fitness 86%
• mortality 74%
• cause 52%
• minority 37%
• learning 30%
• performance 25%
• evaluation 15%
• neural network 13%
• Medical records 13%
• learning method 10%
• health 10%
• clinician 9%
• experiment 6%

Engineering & Materials Science

• Machine learning 64%
• Sampling 61%
• Testing 45%
• Classifiers 35%
• Random forests 26%
• Bayesian networks 25%
• Decision trees 24%
• Support vector machines 20%
• Health 18%
• Exercise equipment 15%
• Neural networks 7%
• Experiments 4%