Getting Started With Machine Learning

 .

Data Exploration
1) Variable identification
2) Cleaning of data
3) Transformation
4) Types of analysis
5) Missing values treatment
6) Outlier treatment

outlier = data point that doesn't seem to fit within the set.

Feature  identification

select the most relevant and appropriate features of your data

The  curse  of  dimensionality

In ML applications, we often have high-dimensional data

Feature  selection  and  feature  extraction

Feature Selection

using only variables or features that are relevant to the problem at hand

    

Feature extraction

focus on combining existing features into new, derived features that better represent the data while also eliminating extra or redundant dimensionality

Principal Component Analysis

sophisticated feature extraction technique

Pearson  correlation  analysis

finding correlation among features

Cleaning  and  preparing  data

--Handling  missing  data

----Missing  categorical  data

----Missing  numerical  data

--Handling  noise

--Handling  outliers

--Transforming  and  normalizing  data

[MLJS][MLJS02]