Machine Learning – Introduction

Machine Learning : it is a system where system will learn how the pattern is going and accordingly we can predict the future information.

There are two types in machine learning:

1. Classification
2. Regression

Classification:

in this category it would be like – true or false. For any condition system has to predict if the output would be true or false. Standard example of ML on this would be an email would be SPAM or NO-SPAM how do we predict that.their are similar examples for it.

ML has multiple algorithms for classification.

Regression:

this is ML for prediction of the future expected value per the given data upon pattern of the input data and sample data to predict the future data on the given sample. Their are different algorithms depending upon the pattern of data and the combination of input data values.

• The important about prediction would be the have input data would be only NUMBERS and prediction of that would be numbers. It will not support other then numbers. It is based on mathematical calculation.
• The calculation part is integrated in the model so it is not necessary to be expert in mathematics, but good to have knowledge of it.
• ML should have only data, it contains two setup of values – row data (Training data) and sample data (Test Data) to predict
• Different model fit works for different type of data and works best for some and some do not work well.
• It could be possible that the model works best today would not work well in future so continuous testing is needed.
• Prediction never be 100% as it is on algorithm it would be mostly good but it would never be 100% accurate.
• Their are multiple ways to validate the prediction – rg. Least mean square root criteria – errors.