Skip to content

Files

Latest commit

chboeychboey
Add files via upload
Feb 27, 2024
32a8f2d · Feb 27, 2024

History

History
81 lines (41 loc) · 3.1 KB

readme.md

File metadata and controls

81 lines (41 loc) · 3.1 KB

PCR Classification and RFS Regression

Running the file

  1. Check if your interpreter is missing any libraries or packages, if so there is a requirements.txt file, you can use the command pip install -r requirements.txt

2a) We have provided you a joblib files for each of our models, please ensure that the TestDatasetExample.xls file is in the parent folder COMP3009_Coursework_2_Code_Group_12, so that you would be able to run FinalTestPCR.py and FinalTestRFS.py.

2b) Ensure that you do not change the path of the joblib files as well! [AdaBoost_Model.joblib] -> For PCR Classification [RandomForest_Regressor_Model] -> For RFS Regression

  1. Once you have ensured that you have adhered to steps 1 to 2b, you can run FinalTestPCR.py and FinalTestRFS.py, Happy Testing!! :D

Classification

  1. If you would like to look at our implementation for PCR classification, it would be under a subfolder PCR_Classification, which contains three file:
    • GridSearch_PCR_Classification.py
    • PCRModelEvaluation.py
    • Preprocessing_PCR_Classification.py
  • GridSearch_PCR_Classification.py

Contains the implementation of GridSearch CV for Hyperparameter Tuning of Classification Models

  • PCRModelEvaluation.py

Contains the analysis of the performance of models such as: Balanced Accuracy, Precision and the Accuracy for Negative and Positive Classes. Apart from that, there's visualisation to capture the number of True Positives, True Negatives, False Positives, False Negatives with Confusion Matrices for each model.

This python file would generate the AdaBoost_Model.joblib file.

  • Preprocessing_PCR_Classification.py

Contains the analysis of Feature Importance using ANNOVA, with dropping rows and without, and the best value of k.

Regression

  1. Implementation for our RFS Regression can be found in the subfolder RFS_Regression, which contains three file:
    • GridSearch_RFS_Regression.py
    • RFSModelEvaluation.py
    • Preprocessing_RFS_Regression.py
  • GridSearch_RFS_Regression.py

Please note that grid search will take some time. When grid search is completed for a model, the model's best parameters and its corresponding Mean Absolute Error will be printed.

  • RFSModelEvaluation.py

When model evaluation for a model is complete, a plot for the model's predicted values against the actual values will be made, followed by another plot which shows the model's predicted values against the residual values. The Mean Absolute Error for the model will also be printed. After all model evaluation has been completed, a joblib file for the best model (RandomForestRegressor) will be created, with the filename of RandomForest_Regressor_Model.joblib

  • Preprocessing_RFS_Regression.py

Two plots will be generated to show feature importance on the dataset using ANOVA f-scores before and after mean imputation.The plot for feature importance after mean imputation will then be used to decide the number of features to be used during feature selection with ANOVA.

Thank you - Group 12