Change Machine Learning Data Type

[eugeneang12]

Hi, currently doing on a project, and am facing issue to changing data Type from "ID" to "Categorical".

Wish the run the machine learning to detect how popular tag will affect percentage of positive review of each games.

Have also uploaded the original dataset. Anyone can help with this, thank you so much!

[c-lopez]

Hello,
There are a couple of questions in here I will do my best to answer. I see what you mean, the Title column is set to ID when it shouldnt be. A couple of thoughts on this
1. Title is detected as an ID because it is unique for all titles, no two games have the same title. My question is the following, why is this required to be part of the model you are trying to build? What insight would Title provide? Speaking of that, which column is your target column, what are you trying to predict?

2. Popular tag. There are 5 tag columns and each one can have a value that exists in a different column. For example Column 1 of game A has 'RPG' as a tag and Column 5 of game B also has 'RPG' as a tag. How can you weight this? Will every column give you an insight or can we limit this to column 1 or 2 only? 

I hope this helps,

Regards