Exact methods for variable selection via mathematical programming

2016 Fiscal Year Final Research Report

• Project/Area Number: 26560165
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Social systems engineering/Safety system
• Research Institution: Tokyo University of Agriculture and Technology
• Principal Investigator: Miyashiro Ryuhei 東京農工大学, 工学(系)研究科(研究院), 准教授 (50376860)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 特徴選択 / 数理工学 / アルゴリズム / 統計 / ＯＲ

Outline of Final Research Achievements

Regression analysis is a method to extract a hidden model from a large number of observations (samples). In this research, we concentrated on constructing algorithms for feature selection (variable selection) problems. Our algorithms are based on mathematical programming, which aims getting true optimal solutions. We have developed several integer-programming based algorithms, and have shown that the proposed algorithms produced better solutions than ones given by heuristics algorithms of previous researches.

Free Research Field

数理計画法