
The work folder is where all of the Julia files involved in the Machine Learning Pipeline Can Be Found.

Programs should be run in the following order.

machine_learning.jl - performs statistical analysis using database.

Program works in Julia v1.1.1.

packages.jl
-----------
Required packages can be found in this file.

machine_learning.jl
-------------------
This program is used to train a predictive model and evaluate its performance.
Machine Learning utilizes .sql files found in SQL files folder, to create a local MySQL database which machine_learning.jl utilizes as input. A local MySQL database can be created by running .sql files using MySQL.

machine_learning.jl utilizes the following files:
- crossval_and_stats.jl: implements functions for n-fold cross-validation and
    analysis of varied training set sizes. Also implements functions for
    calculating prediction statistics and writing them to STDOUT and outfiles.
- feature_normalization.jl: implements a min-max function defined as such: x' = (x - min(x))/(max(x)-min(x)) for feature scaling and normalization.
- hyperparmeter_optimization.jl: Utilizes Scikit-Learn to implement randomizedsearch 3-fold cross validation for optimization of hyperparameters.
- classifiers.jl: implements helper functions to utilize the various different
    machine learning algorithms.
- feature_selection.jl: used to create a dataframe that contains a column for
    each feature used to train the predictive model.
- auc_functions.jl: implements functions from scikit Learn for determining the area under the Receiver Operator characteristic curve
