There is a plethora of information on analytical modeling on the web but I recently had a coworker ask me to teach them. Only if this person new how difficult and numerous hours of studying and after-work tinkering us data scientists do, then the individual would have knew how insulting there question was. Hey ya don't ask a surgeon to teach you how to operate in their free-time.
Note: One key job of a data scientist is collecting and wrangling data into a dataset that can be consumed by an anlytical model. This is where you will spend 80% of your time. Also, I have never been provided a dataset that is ready for modeling.
Modelers guide through the galaxy
please download the following dataset:
#What is our target?
#check for missing variables
#log transformation (not always applicable)
#feature engineering (not always applicable)
#feature selection (sometime we do this multiple times)
#CV
Note: One key job of a data scientist is collecting and wrangling data into a dataset that can be consumed by an anlytical model. This is where you will spend 80% of your time. Also, I have never been provided a dataset that is ready for modeling.
Modelers guide through the galaxy
- Classification, Prediction, Unsupervised
- Missing variables
- Log transformations
- feature engineering
- scratch your head and repeat various steps
- feature selection
- CV
please download the following dataset:
#What is our target?
#check for missing variables
#log transformation (not always applicable)
#feature engineering (not always applicable)
#feature selection (sometime we do this multiple times)
#CV