Supervised Learning —part 1

Machine Learning — 2

Supervised Learning, where data comes with the desired output for a given input.

There are two categories of supervised learning:

1. Regression

• Learn to predict a continuous output value.
• we have input x, we try to learn f(x) = y, then y is continuous output value.
• There is a similarity measure in continuous label space.
• It gives the probability of correctness of result.

Different regression algorithms:

1. Linear Regression

2. Logistic Regression

3. Polynomial Regression

4. Ridge Regression

5. KNN Model

6. Support Vector Machines

7. Gaussian Regression

2. Classification

• we have input x, we try to learn f(x) = y, then y is enumeration value.
• Learn to predict discrete label from a set of pre-defined labels.
• Binary classification: exactly two classes classification.
• Multiclass classification: classification between more than two classes.
• Eg: classifying cat/dog, or the output is 1, 2, or 3.
• two classes are either the same or not.
• There is no similarity measure.

Different classification algorithms:

1. Binary classification
2. Multiclass classification
3. Random forest classifier
4. Stochastic Process
5. Naïve Bayes classifier