SOA/RPC>> Kaggle-House-Prices-Advanced-Regression-Techniques>> 返回
项目作者： sahilsharma884

项目描述 ：
Kaggle Competition: House Prices: Advanced Regression Techniques
高级语言： Jupyter Notebook
项目地址: git://github.com/sahilsharma884/Kaggle-House-Prices-Advanced-Regression-Techniques.git


Kaggle-House-Prices-Advanced-Regression-Techniques

Kaggle Competition (Getting Started): House Prices: Advanced Regression Techniques (Competition Here)

Predict sales prices and practice feature engineering, RFs, and gradient boosting

Software Used: Anaconda, Python 3.8

I have provided requirments.txt (or environment.yml) (if needed).

Description:

  1. +----------------------------------------------------------------------------------------------------------------------------------------+
  2. | .ipynb         | Describe/Operation Performed                                                                                          |
  3. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  4. | Notebook-I:    | Read train and test csv and perform handling missing data.                                                            |
  5. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  6. | Notebook-II:   | EDA and Splitting train into train,cv,and test portion.                                                               |
  7. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  8. | Notebook-III:  | Training RandomForest Regression using all Features and hyperparameters                                               |
  9. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  10. | Notebook-IV:   | Perform Testing stage for above trained model and submitted to Kaggle.                                                |
  11. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  12. | Notebook-V:    | Perform and Training using Feature Selection with RandomForest Regression with best parameters from Notebnook-III     |
  13. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  14. | Notebook-VI:   | Perform Testing stage for above trained model and submitted to Kaggle.                                                |
  15. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  16. | Notebook-VII:  | Training GradientBoosting Regression using all Features and hyperparameters                                           |
  17. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  18. | Notebook-VIII: | Perform Testing stage for above trained model and submitted to Kaggle.                                                |
  19. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  20. | Notebook-IX:   | Perform and Training using Feature Selection with GradientBoosting Regression with best parameters from Notebnook-VII |
  21. +----------------+-----------------------------------------------------------------------------------------------------------------------+
  22. | Notebook-X:    | Perform Testing stage for above trained model and submitted to Kaggle.                                                |
  23. +----------------+-----------------------------------------------------------------------------------------------------------------------+

Result:

  1. +---------------------------------------------------------------+
  2. | Features          | Model    | Test Score (Kaggle Submission) |
  3. +-------------------+----------+--------------------------------+
  4. | All Features      | RF Model | 0.19276                        |
  5. +-------------------+----------+--------------------------------+
  6. | Feature Selection | RF Model | 0.18691                        |
  7. +-------------------+----------+--------------------------------+
  8. | All Feature       | GD Model | 0.20075                        |
  9. +-------------------+----------+--------------------------------+
  10. | Feature Selection | GD Model | 0.15973                        |
  11. +-------------------+----------+--------------------------------+