manual.txt

0. Libraries and download links
    Pandas
    https://pandas.pydata.org/pandas-docs/stable/getting_started/index.html
    Matplotlib
    https://matplotlib.org/users/installing.html
    Numpy
    https://numpy.org/install/
    Scikit-learn
    https://scikit-learn.org/stable/install.html
    Pytorch
    https://pytorch.org

1. Preprocessing
     python preprocess.py 
     preprocess the result


2. Neural Network
    python neural_network.py
    Comment line 680-689 to avoid running experiments with different training sample size to save time.

3. SVM
    Langauge: Python3.7

    Command to run:
    Since hyperparameter tuning takes hours, there is a way to skip it.
    (1) Running without hyperparameter tuning (gamma = 0.01, C = 100)
        python3 svm.py 1
    (2) Running with hyperparameter tuning
        python3 svm.py 0

4. Adaptive Boosted Decision Tree
	python AdaBoostTree.py 
	produce the main result

	python TuningTree.py  
	find the best combination of hyperparameters.

5. Data files
       data/training.csv
       Raw data

       data/trainingSet.csv
       processed training data

       data/testingSet.py
       processed testing data


6. Plots

	plots/plots_raw
	All the features of raw data

	plots/plots_raw_log
	All the features of raw data with log y axis

	plots/plots_preprocessing
	All the features of preprocessed data

	plots/SVM
	SVM results

	plots/AdaBoostTree
	AdaBoostedTree results