Introduction to Machine Learning (3 cr)

Objective

Knowledge and understanding
The students will know some basic concept and some algorithms of machine learning for regression and classification and they will understand how to implement them in Python. They also will know an artificial neural network model for handwritten character recognition and they will be able to implement both the neural network learning and recognition algorithm in Python. To gain a deeper understanding of the basics of neural networks, we don't use the modern frameworks commonly used to implement deep learning.

Skills
The students are able to understand basic machine learning algorithms especially artificial
neural networks and make use of Python for implementation of the algorithms.

Content

Core content level
Introduction to Python programming for implementation of machine learning algorithm.
Introduction to artificial neural networks.
Formal neuron and perceptron.
Simple classification using a single-layer perceptron network.
Multi-layer perceptron network and feed-forward network functions.
Neural network learning based on back-propagation algorithms.
Handwritten character recognition using neural network algorithms.
Supplementary explanation towards deep neural networks.

Additional content
Recognition of handwritten characters written by the participants using the neural network models.

International connections

Course lecturers are Assoc. Prof. Dr. Akihiro Yamashita and Prof. Dr. Hiroyuki Aoki from National Institute of Technology.

Evaluation scale

0-5

Assessment criteria, satisfactory (1)

Daily exercises assigned on the course are worth 50% and both a report and products (i.e.
program source code, sample dataset, experimental results, and so on) about handwritten
character recognition are worth 50%.

Assessment criteria, approved/failed

Daily exercises assigned on the course are worth 50% and both a report and products (i.e.
program source code, sample dataset, experimental results, and so on) about handwritten
character recognition are worth 50%.

Qualifications

Basic mathematics, for example, linear algebra, vector and matrix operations, linear
combination, basic multivariate calculus, and so on. It is preferable to have a basic
programming skill in Python and to be able to use basic data structures and algorithms.

Objective

Knowledge and understanding
The students will know some basic concept and some algorithms of machine learning for regression and classification and they will understand how to implement them in Python. They also will know an artificial neural network model for handwritten character recognition and they will be able to implement both the neural network learning and recognition algorithm in Python. To gain a deeper understanding of the basics of neural networks, we don't use the modern frameworks commonly used to implement deep learning.

Skills
The students are able to understand basic machine learning algorithms especially artificial
neural networks and make use of Python for implementation of the algorithms.

Content

Core content level
Introduction to Python programming for implementation of machine learning algorithm.
Introduction to artificial neural networks.
Formal neuron and perceptron.
Simple classification using a single-layer perceptron network.
Multi-layer perceptron network and feed-forward network functions.
Neural network learning based on back-propagation algorithms.
Handwritten character recognition using neural network algorithms.
Supplementary explanation towards deep neural networks.

Additional content
Recognition of handwritten characters written by the participants using the neural network models.

Qualifications

Basic mathematics, for example, linear algebra, vector and matrix operations, linear
combination, basic multivariate calculus, and so on. It is preferable to have a basic
programming skill in Python and to be able to use basic data structures and algorithms.

Accomplishment methods

Intensive course with hands-on exercises