Information literacy overview
On Air
Subject | Information & Data Science
Introduction to AI and Data Science - Fundamentals and Applications
- AI
- Mathematics
- Impact Evaluation
- Statistics
Through this course, students will learn and understand the significance of studying AI and data science for solving real-world problems and understanding complex phenomena. The course covers foundational mathematical and programming skills, basic approaches and ways of thinking in various related fields, and real-world case studies where these methods are applied.
©University of Tsukuba. Last Update 2025-09-25
Content／学習内容
-
Social and engineering implications of Mathematical, Data science, and AI approaches
This course aims to explain the social and engineering significance of approaches using mathematics, data science, and AI, and to help students understand the usefulness of the fundamental methods from these fields that they will learn throughout the course.
Videos
／学習動画
-
1-1 What is MDA approach and education?
In Japan, developing digital talent capable of appropriately using approaches based on mathematics, data science, and AI has become an urgent priority, and universities across the country are actively addressing this need. This course is designed to teach the fundamentals of such approaches, in line with this national context.
-
1-2 Introduction of Impact Evaluation in Practice
This course introduces the effectiveness of evidence-based policymaking, along with concrete real-world examples.
-
1-3 Example of Data Science for a Social Issue
Using the gender gap in STEM education as a case study, this course introduces an example of how data science can be applied to address social issues.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Programming Basics
In this session we will learn the basic knowledge required to create programs used to run a range of processes on a computer.
Videos
／学習動画
-
2-1 Data types, variables, and basic operations
We will learn the basics of programming, including variables and data types such as strings, lists, tuples, and dictionaries.
-
2-2 Repetitions and branches in programming
We will learn about control structures, which are the mechanisms that determine the order in which instructions in a program are executed.
-
2-3 Functions
We will learn about functions, which allow you to group and execute multiple lines of code together.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Mathematics Basics
We will learn what vectors and matrices are, along with basic operations such as addition and multiplication. We will also explore examples of how vectors and matrices are used in data analysis, including practical calculation examples.
Videos
／学習動画
-
3-1. Basics of Vectors and Matrices
We will learn the definitions of vectors and matrices, as well as concepts such as the dot product of vectors and orthogonality.
-
3-2. Basic Vector and Matrix Operations
We will learn about the basic operations of matrices (addition, subtraction, multiplication, division), as well as the dot product and norms.
-
3-3. How to Use Vectors and Matrices
Matrices are commonly used in real-world applications such as analyzing digital images, text data, and audio signals.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Introduction and Practice of Data Science Part I
You will learn basic methods for visualizing data using graphs. You will also learn how to interpret the information shown in those visualized graphs.
Videos
／学習動画
-
4-1 Data visualization -comparison and change-
You will learn how to visualize data using bar charts, box plots, heat maps, and line graphs, and how to use them to make comparisons and understand changes over time.
-
4-2 Data visualization -distribution and correlation-
You will learn how to visualize data using pie charts, bar strips, histograms, and scatter plots, and how to use them to understand distributions and correlations.
-
4-3 Big-Data and Open-Data
You will learn that different types of data require different visualization methods, and that the variety of available open data has been increasing in recent years.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Introduction to Deep Learning & AI Part A
After explaining what machine learning and deep learning are, we will introduce the basics of unsupervised learning and show programming examples to help you practice it.
Videos
／学習動画
-
5-1 What is Deep Learning & Machine Learning?
We will explain what machine learning and deep learning do, what kinds of problems they can be applied to, and what types of models are used.
-
5-2 What is Unsupervised learning?
As key examples of unsupervised learning, we will explain clustering and probability density estimation.
-
5-3 Exercise of Unsupervised learning
As an introduction to practicing unsupervised learning, we will present example calculations using Python.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Introduction to Deep Learning & AI Part B
As a foundation for supervised learning, we will explain the types of models and the data required. Then, we will introduce specific models—Support Vector Machines (SVM) and Neural Networks (NN)—along with example Python code to demonstrate how they work.
Videos
／学習動画
-
6-1 What is supervised learning?
To build a supervised learning model, you need to provide labeled (correct answer) data as input. To avoid overfitting, techniques such as cross-validation and regularization are also effective.
-
6-2 Detailed models of supervised learning in Machine Learning
As a specific example of a supervised learning model, we will explain Support Vector Machines (SVMs) while showing Python code.
-
6-3 Detailed models of supervised learning in Deep Learning
As a specific example of a supervised learning model, we will explain Newral Network models while showing Python code.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Impact Evaluation in Practice
As a continuation of the deep learning section, we will first explain generative AI. After that, we will introduce practical impact evaluation by explaining its basic concepts and how to carry out the evaluation.
Videos
／学習動画
-
7-1 How to use Generative AI?
Based on the University of Tsukuba’s guidelines for the use of generative AI, this part explains important considerations and key points to understand when using generative AI.
-
7-2 What is Impact Evaluation?
Impact evaluation involves two key processes: monitoring and evaluation. When considering the effect (causal impact) of a program on outcomes, it is necessary to think about the counterfactual—what would have happened without the program.
-
7-3 How to proceed Impact Evaluation?
Impact evaluation includes four preparatory steps: developing a theory of change, defining evaluation questions, selecting indicators, and assessing costs. Among these, building a theory of change is especially important—it involves creating a causal chain that clearly maps out the cause-and-effect relationships from the start to the end of the program.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Introduction and Practice of Data Science Part II
This section explains the basic statistical concepts needed to conduct an impact evaluation.
Videos
／学習動画
-
8-1 Sampling and Statistical Inference
As a foundation of data science, this section explains samples and statistical inference.
-
8-2 Hypothesis test
Hypothesis testing is a statistical method used to determine whether a policy has an impact. The evaluation is carried out by following appropriate steps.
-
8-3 Regression Analysis
We will explain regression analysis by first discussing parameter estimation methods and multicollinearity between explanatory variables. Then, using calculation examples, we will demonstrate how to perform regression analysis and how to interpret the results.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Quantitative evaluation of measures by Causal Inference
We will explain what causal inference is and the basic ideas behind identifying causal effects. Then, we will introduce specific evaluation methods, focusing on simple and practical approaches.
Videos
／学習動画
-
9-1 Basic logic of causal inference (impact evaluation)
We explain the basic concept of identifying causal effects by focusing on counterfactuals. Because confounding factors can distort causal relationships, specific methods are needed to properly identify causal effects.
-
9-2 Evaluation by causal inference (1) (regression analysis, stratification, matching)
In policy evaluation using causal inference, we explain relatively simple methods such as multiple regression analysis, stratification, and matching, using calculation examples to illustrate each approach.
-
Evaluation by causal inference (2) (Difference in Difference)
In policy evaluation using causal inference, the difference-in-differences (DiD) method is an effective approach when time-series data are available. This method will be explained using a calculation example.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
-
Examples of AI and Data Science Deployment in the Real World
This session introduces two real-world case studies where AI and data science approaches have been applied. We will also reflect on the content of the course and explain once again the key skills and knowledge that students are expected to gain through this lecture.
Videos
／学習動画
-
Review of the class
This session will explain the key concepts that students were expected to understand through this course, as well as the areas we hope they will continue to develop further in the future.
-
The 2011 Tohoku-Oki earthquake is not over yet._x000B_ Investigation of Postseismic Deformation after the Tohoku-Oki Earthquake through High-Density Continuous GPS Observations and Data Science
This presentation has a case study that used data science to analyze postseismic deformation data to uncover the mechanisms causing these strange crustal deformations and make future predictions.
-
Spatio-Temporal Analysis for Understanding the Traffic Demand After the 2016 Kumamoto Earthquake
This lecture introduces a case study in which a data science approach was used to reveal that the estimated population data during the disaster recovery phase reflects both regular-purpose and recovery-related movements.
Lecturers
／講師
-
Junji Urata
University of Tsukuba
-
Mikio Tobita
University of Tsukuba
-
Staff／スタッフ
-
Junji UrataUniversity of Tsukuba -
Mikio TobitaUniversity of Tsukuba
Competency／コンピテンシー
Course Objectives
The goal is to acquire foundational knowledge for practical applications in AI and data science, and to become able to conduct basic analyses using approaches from these fields.
Learning Outcomes
Students will gain fundamental knowledge for practical applications in the fields of AI and data science, and develop the ability to carry out basic analyses using approaches from these fields.
Contact／お問合せ先
Prof. Urata’s lab: https://uratalab.net/contact/
