Learning Path of R

r

#1

Basically as of now i am aware that there are three stages of modelling.

  1. Pre modelling stage
  2. Modelling Stage
  3. Post Modelling Stage

So my question here is that is it necessary to learn the packages in the order of the modelling stages or can be learnt in any random order.

Eg:

  1. For Pre modelling we have ggplot2
  2. For modelling we have RandomForest
    3.For Post Modelling we have RegTest

So do i have to learn in modelling stage order or can be learnt in any order?

Thanks in advance


#2

Hi @harshabeesai1026

This is my point of view EDA first that is 80% of the effort of a modelling project. You call it pre modelling, your knowledge of models will give you hints at this stage.

Hope it helps a little

Alain


#3

Thank you very much for the reply. But it would be helpful if you can elaborate the answer you want to tell me.


#4

Hi @harshabeesai1026,

As @Lesaffrea has aptly point, the pre-modelling stage or EDA (Exploratory Data Analysis) is the most important step in Data science. If you don’t know the data, you can’t solve the problem completely. So you should focus on it as your first step. The rest could be done in the order you described.

I would like to add that instead of concentrating on the packages, you should first try to understand the underlying concepts. These help you in a long run, as packages get deprecated easily, but ideas don’t.


#5

@harshabeesai1026

The philosophy behind following any learning path is to provide structured path to the end user. Going in sequential manner would help tremendously.

While nothing stops you from changing the order, it will unnecessarily increase the confusion in the learning process. I am not sure why would you want to do that.

Regards,
Kunal


#6

Thank you sir for the detailed reply.