##CS229

###Collecting and Cleaning Data 1.Download the sqlite database from https://www.kaggle.com/kvnchn/notebook399529d6ab/data

2.Run the cells in Data.ipynb

• Extracts associates match and team attribute data from the above database
• Drops sparse rows and unnecessary columns
• Writes processed data to CSV files

3.Run the cells in scrape_team_attributes.ipynb

• First, you may need to download historical betting odds from football-data.co.uk & name the CSVs appropriately
• Scrapes team attributes from SOFIFA.com for recent EPL seasons
• Associates them with historical betting odds and match results from football-data.co.uk

4.Run the cells in fill_normalize_PCA.ipynb

• Fills missing betting odds using average of present betting odds
• Fills missing team attribute data using K-nearest neighbors
• Normalizes data
• Reduces dimensionality with PCA

5.visualize_data.ipynb

• For creating charts, tables, and visuals

####Training and Testing 1.Neural Net:

• Train/Test: train.py
• Test on Seperate dataset: predict.py "weights file name" NN
• Class definition: Network.py

2.Logistic Regression:

• Train/Test: lr_pca_test.py
• Test on Seperate dataset: predict.py "weights filename" LR
• Class definition: log_reg_network.py

3.Decision Tree:

• Train/Test: train_DT.py

4.Gradient Boosting:

• Train/Test: train_GB.py

5.Naive Bayes:

• Train/Test: train_NB.py

6.Random Forest:

• Train/Test: train_RFC.py

7.SVM:

• Train/Test: train_SVM.py

8.AdaBoost:

• Train/Test: train_adaBoost.py

9.Q-learning:

• Train: betRL.py "Model type"
• Test: AutoBetter.py "Model type"
• environment definition: env/env.py