I. Introduction
1.1 - Welcome
- Machine Learning is used everywhere: automation, post mail, brain study, Natural language processing (NLP), computer vision, etc… It is one of the top fields for IT skills.
1.2 - What is Machine Learning
- Machine learning definition:
- Arthur Samuel (1959). Machine Learning: Field of study that gives computers the ability to learn without being explicitly programmed.
- Tom Mitchell (1998) Well-posed Learning Problem: A computer program is said to learn from experience E with respect to some task T and some performance measure P, if its performance on T, as measured by P, improves with experience E.
- Main types of algorithms:
- Supervised learning
- Unsupervised learning
- Also: Reinforcement learning, recommender systems.
1.3 - Supervised Learning
- Regression : trying to generate a continuous output value from the input sample.
- Classification: trying to assign a clearly define discrete value to the inpu sample.
- For classification problems we may have more than one feature to classify on.
- We can even deal with an infinite number of features with the Support Vector machine algorithm.
1.4 - Unsupervised Learning
- Clustering algorithm: used to separate the dataset samples into different clusters.
- With unsupervised learning, the idea is to automatically find structure into the dataset.
- Can be used to separate voices from two microphones for instance. Or separte voice from music (still with two microphones).