Machine Learning

Types of machine-Learning

Supervised learning

• Classification
  Regression

Unsupervised learning

• Clustering
• Anomaly detection

Other types

Supervised Learning

Input: Dataset of training examples

• Input data (variables/features)
• Desired output (label/target/class)

Result: predictive model

• Input data -> label

Find underlying, predictive relationships
Important: Data has to be labeled (with desired outcome)

Supervised Learning: classification  

Data points are divided into classes    (label = class)

Model predicts what class a new data point should belong to

Example algorithms:
Logistic Regression, k-NN, Neural Networks, Decision Trees

Example Applications:

• Churn prediction
• Image recognition
• Medical diagnosis

Supervised Learning: Regression

Given a historical list of instances with numerical values…     (label = numerical value)
Predict the numerical value of a new instance

Example algorithms:
Linear Regression, k-NN, Neural Networks

Example applications:

• Stock market prediction
• House price prediction
• Sales prediction

What is a good model

Not good: fit the given data perfectly (overfitting)
a model must generalize well