notes.txt

https://arxiv.org/pdf/1710.02824.pdf
https://www.sportsbookreviewsonline.com/scoresoddsarchives/nba/nbaoddsarchives.htm
https://www.youtube.com/watch?v=wQ8BIBpya2k
Random Forrest
https://www.analyticsvidhya.com/blog/2015/09/random-forest-algorithm-multiple-challenges/?utm_source=blog&utm_medium=understandingsupportvectormachinearticle
naive bayes
Support Vector Machine
LSTM
Recurrent
Support Vector Regression

recurrent NN


https://www.youtube.com/watch?v=BqgTU7_cBnk

Hyperparameter tuning
https://www.youtube.com/watch?v=vvC15l4CY1Q

FIX ODDS-DATA
1-24-16 Raptors.....OKC

Miles travled since last game?
Days since last game

https://klane.github.io/databall
https://www.covers.com
https://www.oddsportal.com

Leaky ReLu for most
try elu and selu

# xgb.plot_tree(model)
# plt.show()
# plt.savefig("graph.pdf")

look into tensorboard/ callbacks. Improve NN arc. See what data is helpful with a confusion matrix.
Normalize data?
Improve NN performance
https://machinelearningmastery.com/improve-deep-learning-performance/
https://www.tensorflow.org/tutorials/keras/keras_tuner
https://www.youtube.com/watch?v=k7KfYXXrOj0
https://keras.io/api/layers/
https://medium.com/@Mandysidana/machine-learning-types-of-classification-9497bd4f2e14


More Data:
    Depth 2 - 67.38399999999999
    Depth 3 - 67.34200000000001
    Depth 5 - 67.12200000000001
    Depth 7 - 66.62600000000002

    https://www.gamblingsites.org/sports-betting/kelly-criterion/

https://cs229.stanford.edu/proj2018/report/3.pdf
    Neural Network: We use four fully-connected layers, reducing the input of dimension 1524 to size 500, then 100,
    then 20, then finally a 1-dimensional output that predicts the Over- Under of the desired game (see Figure 2).
    We used a weight decay parameter of 1 to regularize the weights of the network,
    and a learning rate of 10−6 over 5000 epochs, using Gradient Descent.
    This allowed the network to roughly train to convergence (see Figure 5)

    When we were training the models, we found while we had data available starting from the 2007 − 2008 season,
    we achieved the highest levels of validation accuracy when the training dataset started from the 2012 − 2013 season
    rather than from the 2007 − 2008 season.
    This suggests that the earlier seasons are not very representative of the validation season
    (which was the 2016 − 2017 season) possibly because of the changing nature of the NBA