Day-21

Day 21 of #100DaysOfCode

Task

1. Intermediate Machine Learning Course
2. House price prediction

Resources

• Kaggle's Intermediate Machine Learning course

• • Lesson 4: Pipelines - My Notebook
• • Lesson 5: Cross Validation - My Notebook
• • Lesson 6: XGBoost - My Notebook
• • Lesson 7: Data Leakage - My Notebook

• Kaggle House price prediction Challenge

• • My Notebook 1
• • My Notebook 2
• • My Notebook 3

Topics I have learnt

1. Intermediate Machine Learning Course

• Using Pipelines to write structural code
• Cross validation using cross_val_score, (use when data is less, takes more time than train_test_split)
• XGBoost using XtremeGradientRegressor aka XGBRegressor
• Data Leakage, how to handle data leakage

2. House price prediction

• One by Imputing missing values, doing Ordinal Encoding using RandomForestRegressor.
• One mostly same as above but used more features
• One using Cross Validation and Pipelines with RandomForestRegressor.

Software used

• Jupyter Notebook
• Python 3.10.2
• Numpy 1.22.4
• pandas 1.4.2
• Matplotlib 3.5.2
• Seaborn 0.11.2
• scikit-learn 1.1.1
• XGBoost 1.6.1

My Notebooks

• L4 - Pipelines.ipynb
• L5 - Cross_validation.ipynb
• L6 - XGBoost.ipynb
• L7 - Data_leakage.ipynb
• House_price_prediction_4.ipynb
• House_price_prediction_5.ipynb
• House_price_prediction_6.ipynb

Conclusion:

Today I learned how to use pipelines to write cleaner code, cross validation using cross_val_score, XGBRegressor and how to handle data(Target) leakage and train_test_contamination. Also, did house price prediction using cross validation.