Skip to content

ecdison/Group9_MachineLearningHC

Repository files navigation

Group9_MachineLearningHC

Model Ensemble

Introduction

model ensemble is a method to get a better result from the voting of a group of model, therefore we must ensure the models are using the same training and testing set.

Step

  1. Ensure the code in the branch is the latest version (we had updated the code and dataset)

  2. run the "HCTech.ipynb" code until completing the following block

X_train.to_csv('train_binary_x.csv', index=False)
y_train.to_csv('train_binary_y.csv', index=False)
X_test.to_csv('test_binary_x.csv', index=False)
y_test.to_csv('test_binary_y.csv', index=False)

*This block save the data to the current directory, you can change its path.

  1. Use 'train_binary_x.csv' and 'train_binary_y.csv' to train the model (decision tree, random forest, MLP, perceptron, DNN or anything) *You can adjust the thing used to change (ex: using SMOTE) but ensure to start with these specific file.

  2. Use 'test_binary_x.csv' to test the model, DO NOT use 'test_binary_y.csv' to evaluate it.

  3. Output the answer as the following format.

*Do not change the order of the 'test_binary_x.csv'

index, score
0, 1.8863079276343342e-06
1, 1.0374536032031756e-05
2, 5.440356289909687e-06
3, 0.7755006551742554
4, 1.4836804439255502e-05
5, 1.5318420992116444e-05
6, 3.7484101085283328e-06
7, 2.6681723284127656e-06
8, 6.323603156488389e-05
9, 3.993187874584692e-06
10, 3.346433231854462e-06
11, 3.931617357011419e-06

The first line should be written as 'index, score\n' The first column should be the index of the testing data. The seconf column should be ==the probility of having a birth (predict 1)==

  1. Save the answer as the format of "result-{method}-{accuracy_of_training_data}.txt" ex: result-RandomForest-0.954321.txt, result-DNN_SMOTE-0.954321.txt

*accuracy_of_training_data should be with at least six decimal places

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •