Generating Synthetic Samples to Improve Small Sample Learning with Mixed Numerical and Categorical Attributes

Business & Economics

• Small Sample 100%
• Sign Test 31%
• Student-t Distribution 28%
• Sample Size 27%
• Back-propagation Neural Network 27%
• Support Vector Regression 27%
• Machine Learning 26%
• General Population 25%
• Membership Function 24%
• Network Model 20%
• Regression Model 16%
• Industry 7%

Engineering & Materials Science

• Support vector regression 59%
• Membership functions 52%
• Backpropagation 47%
• Students 37%
• Machine learning 37%
• Neural networks 30%
• Industry 22%

Social Sciences

• learning 34%
• neural network 20%
• efficiency 12%
• regression 10%
• cause 9%
• industry 9%
• student 4%