Supervised and Unsupervised Machine Learning

What is supervised Machine Learning ?
Take input as data  and generates model that predicts response to new data.

The data can be :
1)Classification : Samples belonging from different classes the algorithm learns from already labeled data to predict the class of unlabled data. It can have values 0 and 1 .An example of classification problem will be to predict weather tumor is Malignant or benign .
2)Regression :  If desired output has one or more contiues value the task is called regression.Miles per gallon for a car.


What is unsupervised Machine Learning ?
Input data consist of set of vectors with out any corresponding target value.Goal is to determine set of similar examples within set of data. Or to determine distribution of data.called density estimation or reducing the dimensionality of data.

* taken from http://bioinformatics.oxfordjournals.org/content/24/6/783/F1.expansion.html

Steps for Supervised machine Learning :
1.Prepare data
2.Choose ALgorithm
3.Fit Model.
4.Validation Model
5.Use it for predication

Various learning algorithm :
Classification Trees   
Regression Trees   
Discriminant Analysis (classification)
K-Nearest Neighbors (classification)   
Naive Bayes (classification)   
Classification or Regression Ensembles 
Classification or Regression Ensembles in Parallel