We know the three types of problem we handle. See Below

• Classification
• Regression
• Optimization
• Unsupervised Learning

Based on the problem type we select Machine Learning Model. For example, when we have to deliver the client in terms numbers and probabilities, we have to choose Naive Bayes Algorithm, and the Decision Tree is for rules.

And Every Model has three components. Those are

• Representation
• Evaluation
• Optimization

Types of Machine Learning Algorithm

• Supervised
• Unsupervised
• Recommended Systems

We know the types of Problems

• Classification
• Regression
• Optimization
• Unsupervised Machine Learning

In Above types,

• Classification & Regresssion comes under Supervised ML
• Optimization comes under Recommended Systems
• Clustering comes under Unsupervised Machine Learning

Regression

When we have to predict continuous variable, it comes under regression.

Classification

When we have to classify something, example, spam or no spam, this type of Algorithm comes under Classification.

Clustering

Social Network Analysis comes under clustering

Collaborative Filtering

Netflix Recommendation comes under Collaborative Filtering.

Supervised ML ( Regression & Classification )

In Supervised ML,

• We have a Target Variable through which we train our Model.
• Predict Class or Label Value ( Regresssion or Classification)
• Examples : Decision Tree, Naive Bayes, Support Vector Machine, Nueral Networks

  Unsupervised ML

• We dont have a Target Variable.
• Here we determine data Patterns and Group
• Example : - K - Means, genetic Algorithm, Clustering

Develop ML Model

• Divide the dataset into Train and Test Data.
• Develop the Model onto Train Data.
• Test the Model onto Test Data and Measure the accuracy.

Applications

• Speech Recognition
• Natural Language Processing
• Computer Vision
• Aritifical Intelligence
• Medical Outcome Analysis
• Robot Control
• Computational biology