Introduction

Definition

There are several definition of Machile learnings:

  • Arthur Samuel (1959)

    • The field of study that gives computers the ability to learn without being explicitly programmed.

  • Tom Mitchel (1998)

    • A computer program is said to learn from experience E with respect to some class of task T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E

Types of ML

In ML, there are three algorithms:

../_images/ml_algorithms.png

Supervised learning

Point

  • “Right answers” are given

Types

  • Regression (Link)

  • Classification.

Regression

  • Map input variables to some continuous functions to predict results within a continuous output

  • Example: Housing price prediction

../_images/regression_ex.png

Classification

  • Map input variables into discrete categories to predict results within a discrete output

  • Example: Breast cancer or Test grade (A, B, C, D, F)

../_images/classification_ex1.png
../_images/classification_ex2.png

Unsupervised learning

Point

  • Allow us to approach problems with little or no idea what our results should look like

  • Derive the structure from data where we don’t necessarily know the effect of the variables

  • No feedback based on the prediction results

Types

  • Clustering

  • Non-clustering

Clustering

  • Find groups with patterns being close to each other

  • Example: Google news clustering

../_images/clustering_ex.png

Non-clustering

  • Find the structure in a chaotic environment

  • Example: Cocktail party problem

../_images/non-clustering_ex.png

Reinforcement learning

Reinforcement learning (RL) is an area of machine learning concerned with how software agents ought to take actions in an environment so as to maximize some notion of cumulative reward (Ref.: Wikipedia).

../_images/reinforcement_learning.png

출처