Support vector regression in high dimension: application and comparison with parametric methods

• Grant number: 19/12774-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: August 01, 2019
• End date: July 31, 2020
• Field of knowledge: Physical Sciences and Mathematics - Probability and Statistics - Applied Probability and Statistics
• Principal Investigator: Chang Chiann
• Grantee: Danilo Vieira Silva
• Host Institution: Instituto de Matemática e Estatística (IME). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

In this work, we will study the theoretical methodology of support vector regression (SVR), aimed at analyzing the data sets with high dimension. Our objective is to compare the performance of this technique with conventional parametric techniques. The parametric methods to be compared are: Ridge regression and Lasso. These parametric methods are most commonly used in high-dimensional problems. In order to illustrate the methods studied in this project, we will use a database of Kaggle (a platform for Data Science competitions) called "House Prices: Advanced Regression Techniques", which consists of the sale price of houses in dollars with 79 explaining the aspects of the houses.