Charity Funding Success Predictor
(Deep Machine Learning Demonstration)

Overview:

This project utilizes a TensorFlow Keras neural network model to predict the successfulness funding various charities for Alphabet Soup. It illustrates how prediction performance can be improved with various data preprocessing techniques and also careful model parameter optimization.

To examine the data, code files, and final report, their locations are indicated in the tree below.
(Note: the notebooks and h5 files should be run in Google Colaboratory.)

Files/Folders:

• "Code" (this folder contains the jupyter notebooks created for this assignment)
  
  • "AlphabetSoupCharity.ipynb" (initial model setup and performance test)
    
  • "AlphabetSoupCharity.h5" (HDF5 version of the model)
    
  • "AlphabetSoupCharity_Optimization.ipynb" (optimized model setup and performance test)
    
  • "AlphabetSoupCharity_Optimization.h5" (HDF5 version of the model)
    
• "Data" (this folder contains the input data file)
  
  • "charity_data.csv" (file containing the training/testing data)
    
• "Final Report" (this folder contains the final report)
  
  • "Deep Learning Final Report.pdf" (read to find out the details of the model creation, optimization, and results)
    

(Please do not delete, move, rename, or alter!)

Sample Code:

Reading Data:

Optimizing Data:

Model Creation:

Model Performance Evaluation: