Methodology of Big Data Analytics

It is slightly varies from conventional approach
It begins with data to describe the reaction
This model forecasts the response behaviour
It produces data in a manner that are to be utilized by a statistical model
It considers certain assumptions like independence, randomization and normality
It represents data , the cleaning process will be taken care by machine learning model
There is no individual method, once the problem is addressed; a research phase is needed for design methodology.
The key important job of big data analytics is statistical modeling which means supervised and un supervised classification or regression problem
Before modelling, data cleaning and pre-processing is done
Before implementing a model care should be taken in evaluating different models
After evaluation the results must be reported