Started at 30 days of ML by Kaggle
Week 2 - I've been working on the Machine Learning Intermediate Kaggle course as part of the 30 Days Of ML 

So far this week I have learned more about: 

  • How to treat missing values and categorical values by using Strategies such as Imputation and Encoding ( Ordinal and OneHot) 
  • The reasons why it's important to make use of Pipelines for Data processing and training. For example, one of the exercises was about writing your own Random Forest Tree model and perform predictions based on the trained model, then found that I had to repeat all the processing steps again by hand, and Boom!!! .... Here it comes the pipelines to save you :) 
  • Last but not least learned about XGBoost which is something that I heard about long back but never explored and finally got to see why it's so important in Regression and how by using this technique you could achieve really good accuracy.