KA240801 Sparse Estimation

Hiroshima University Syllabus

Japanese

Academic Year 2025Year School/Graduate School School of Informatics and Data Science

Lecture Code KA240801 Subject Classification Specialized Education

Subject Name スパース推定

Subject Name
（Katakana）スパーススイテイ

Subject Name in
English Sparse Estimation

Instructor NAGAHARA MASAAKI

Instructor
(Katakana) ナガハラ　マサアキ

Campus Higashi-Hiroshima Semester/Term 3rd-Year, First Semester, 1Term

Days, Periods, and Classrooms (1T) Tues9-10,Weds9-10：ENG 219

Lesson Style Lecture Lesson Style
(More Details) Face-to-face, Online (on-demand)

Credits 2.0 Class Hours/Week 4 Language of Instruction J : Japanese

Course Level 2 : Undergraduate Low-Intermediate

Course Area（Area） 25 : Science and Technology

Course Area（Discipline） 02 : Information Science

Eligible Students

Keywords Machine Learning, Feature Selection, Data Analysis, Regression, Sparse Modeling, Convex Optimization, Combinatorial Optimization, Optimal Control, Energy Saving / Energy Conservation

Special Subject for Teacher Education 　 Special Subject 　

Class Status
within Educational
Program
(Applicable only to targeted subjects for undergraduate students)
Criterion referenced
Evaluation
(Applicable only to targeted subjects for undergraduate students) Computer Science Program
（Abilities and Skills）
・A. Information infrastructure development technology, information processing technology, technology that analyzes data and creates new added value.

Data Science Program
（Knowledge and Understanding）
・D1. Knowledge and ability to understand the theoretical framework of statistics and data analysis and to analyze qualitative/quantitative information of big data accurately and efficiently.
（Abilities and Skills）
・A. Information infrastructure development technology, information processing technology, technology that analyzes data and creates new added value.
・B. Ability to identify new problems independently and solve them through quantitative and logical thinking based on data, multifaceted perspectives, and advanced information processing and analysis.

Intelligence Science Program
（Abilities and Skills）
・A. Information infrastructure development technology, information processing technology, technology that analyzes data and creates new added value.
・B. Ability to identify new problems independently and solve them through quantitative and logical thinking based on data, multifaceted perspectives, and advanced information processing and analysis.

Class Objectives
/Class Outline In large-scale machine learning models, an increase in the number of features leads to a decrease in model explainability and interpretability. Feature selection is an effective technique for such cases. In this lecture, we will study sparse modeling techniques, which are a subset of feature selection methods. We will learn how to build machine learning models, perform feature selection, and use convex optimization techniques for sparse modeling in Python through actual programming, while also thoroughly studying the underlying theory.

Class Schedule Lesson 1: Lecture Overview, Introduction to Sparse Modeling
Lesson 2: Feature Selection Practice Using Kaggle Competition Data
Lesson 3: Group Testing
Lesson 4: Group Testing Practice Using Python
Lesson 5: Sparse Modeling Solution Methods
Lesson 6: Sparse Modeling Solution Methods: Practice Using Python CVXPY
Lesson 7: Regression and Sparse Modeling
Lesson 8: Regression and Sparse Modeling: Practice Using Python CVXPY
Lesson 9: Introduction to Convex Optimization for Sparse Modeling
Lesson 10: Gradient Methods and Subgradient Methods: Practice Using Python
Lesson 11: Convex Optimization for Sparse Estimation (1)
Lesson 12: Convex Optimization for Sparse Estimation (1): Practice Using Python
Lesson 13: Convex Optimization for Sparse Estimation (2)
Lesson 14: Convex Optimization for Sparse Estimation (2): Practice Using Python
Lesson 15: Sparse Modeling: Introduction to Real Applications

Students are required to submit reports detailing the outcomes of their practical exercises. Additionally, a final examination will be held.

Text/Reference
Books,etc. M. Nagahara, Sparse Modeling, Corona, 2027.
https://www.coronasha.co.jp/np/isbn/9784339032222/

PC or AV used in
Class,etc. Text, Handouts, Visual Materials

(More Details) Please be sure to bring your own PC. Please set up an environment where Python can run in advance. We recommend using "Google Colaboratory," which is available in a browser. Please check it in advance. https://colab.research.google.com/

Learning techniques to be incorporated Discussions

Suggestions on
Preparation and
Review On-demand lecture videos will be provided. Please be sure to watch these videos each time and review the content. Following the content of the videos, you will engage in programming exercises during the lecture.

Requirements

Grading Method Grades will be evaluated based on the content of the reports and the final exam.

Practical Experience

Summary of Practical Experience and Class Contents based on it

Message Sparse modeling, or sparse estimation, is an important concept in machine learning and data analysis, and a necessary technique for building machine learning models. Through this lecture, you can thoroughly learn how to use sparse modeling and the theory behind it. This knowledge will surely be useful in the future!

Other

Please fill in the class improvement questionnaire which is carried out on all classes.
Instructors will reflect on your feedback and utilize the information for improving their teaching.

Academic Year	2025Year	School/Graduate School	School of Informatics and Data Science
Lecture Code	KA240801	Subject Classification	Specialized Education
Subject Name	スパース推定
Subject Name （Katakana）	スパーススイテイ
Subject Name in English	Sparse Estimation
Instructor	NAGAHARA MASAAKI
Instructor (Katakana)	ナガハラ　マサアキ
Campus	Higashi-Hiroshima	Semester/Term	3rd-Year, First Semester, 1Term
Days, Periods, and Classrooms	(1T) Tues9-10,Weds9-10：ENG 219
Lesson Style	Lecture	Lesson Style (More Details)	Face-to-face, Online (on-demand)

Credits	2.0	Class Hours/Week	4	Language of Instruction	J : Japanese
Course Level	2 : Undergraduate Low-Intermediate
Course Area（Area）	25 : Science and Technology
Course Area（Discipline）	02 : Information Science
Eligible Students
Keywords	Machine Learning, Feature Selection, Data Analysis, Regression, Sparse Modeling, Convex Optimization, Combinatorial Optimization, Optimal Control, Energy Saving / Energy Conservation
Special Subject for Teacher Education		Special Subject
Class Status within Educational Program (Applicable only to targeted subjects for undergraduate students)
Criterion referenced Evaluation (Applicable only to targeted subjects for undergraduate students)	Computer Science Program （Abilities and Skills）・A. Information infrastructure development technology, information processing technology, technology that analyzes data and creates new added value. Data Science Program （Knowledge and Understanding）・D1. Knowledge and ability to understand the theoretical framework of statistics and data analysis and to analyze qualitative/quantitative information of big data accurately and efficiently. （Abilities and Skills）・A. Information infrastructure development technology, information processing technology, technology that analyzes data and creates new added value. ・B. Ability to identify new problems independently and solve them through quantitative and logical thinking based on data, multifaceted perspectives, and advanced information processing and analysis. Intelligence Science Program （Abilities and Skills）・A. Information infrastructure development technology, information processing technology, technology that analyzes data and creates new added value. ・B. Ability to identify new problems independently and solve them through quantitative and logical thinking based on data, multifaceted perspectives, and advanced information processing and analysis.
Class Objectives /Class Outline	In large-scale machine learning models, an increase in the number of features leads to a decrease in model explainability and interpretability. Feature selection is an effective technique for such cases. In this lecture, we will study sparse modeling techniques, which are a subset of feature selection methods. We will learn how to build machine learning models, perform feature selection, and use convex optimization techniques for sparse modeling in Python through actual programming, while also thoroughly studying the underlying theory.
Class Schedule	Lesson 1: Lecture Overview, Introduction to Sparse Modeling Lesson 2: Feature Selection Practice Using Kaggle Competition Data Lesson 3: Group Testing Lesson 4: Group Testing Practice Using Python Lesson 5: Sparse Modeling Solution Methods Lesson 6: Sparse Modeling Solution Methods: Practice Using Python CVXPY Lesson 7: Regression and Sparse Modeling Lesson 8: Regression and Sparse Modeling: Practice Using Python CVXPY Lesson 9: Introduction to Convex Optimization for Sparse Modeling Lesson 10: Gradient Methods and Subgradient Methods: Practice Using Python Lesson 11: Convex Optimization for Sparse Estimation (1) Lesson 12: Convex Optimization for Sparse Estimation (1): Practice Using Python Lesson 13: Convex Optimization for Sparse Estimation (2) Lesson 14: Convex Optimization for Sparse Estimation (2): Practice Using Python Lesson 15: Sparse Modeling: Introduction to Real Applications Students are required to submit reports detailing the outcomes of their practical exercises. Additionally, a final examination will be held.
Text/Reference Books,etc.	M. Nagahara, Sparse Modeling, Corona, 2027. https://www.coronasha.co.jp/np/isbn/9784339032222/
PC or AV used in Class,etc.	Text, Handouts, Visual Materials
(More Details)	Please be sure to bring your own PC. Please set up an environment where Python can run in advance. We recommend using "Google Colaboratory," which is available in a browser. Please check it in advance. https://colab.research.google.com/
Learning techniques to be incorporated	Discussions
Suggestions on Preparation and Review	On-demand lecture videos will be provided. Please be sure to watch these videos each time and review the content. Following the content of the videos, you will engage in programming exercises during the lecture.
Requirements
Grading Method	Grades will be evaluated based on the content of the reports and the final exam.
Practical Experience
Summary of Practical Experience and Class Contents based on it
Message	Sparse modeling, or sparse estimation, is an important concept in machine learning and data analysis, and a necessary technique for building machine learning models. Through this lecture, you can thoroughly learn how to use sparse modeling and the theory behind it. This knowledge will surely be useful in the future!
Other
Please fill in the class improvement questionnaire which is carried out on all classes. Instructors will reflect on your feedback and utilize the information for improving their teaching.

Back to syllabus main page