Machine Learning

Front 1

Supervised learning goal

Back 1

to learn a mapping from input x to output y given set of input-output labeled dataset

Hint 1

labeled

Front 2

Unsupervised Learning goal

Back 2

Find interesting pattern in an unlabeled dataset

Front 3

Classification

Back 3

Learn mapping of input x into ouputs class y.

Front 4

Binary classification

Back 4

Classification with number of output class equal to 2

Front 5

Multiclass classification

Back 5

Classification where number of output class is more than 2

Front 6

Multilabel classification

Back 6

Classification where an input can be classified into several output class

Front 7

Regression

Back 7

Mapping of input x into y where the response variable is continous

Front 8

Example of unsupervised learning

Back 8

1. Discovering cluster

2. Discovering latent factor

3. Discovering graph structure

Front 9

PCA

Back 9

Principal Component Analysis. os an unsupervised learning which does dimensionality reduction.

Front 10

Parametric model

Back 10

ML model which has fixed number of parameter

Front 11

Non parametric model

Back 11

ML model which has growing number of parameter as the training set grows

Front 12

Overfitting

Back 12

Model behavior which perform good in training set but doesn't capture generalized behavior of the training set which then perform bad during testing

Front 13

Underfit

Back 13

Model behavior that produce high error against the real result

Front 14

No free lunch theorem

Back 14

all models are wrong but some are useful