/Neural_Network_Charity_Analysis

Primary LanguageJupyter Notebook

Neural Network Charity Analysis

Overview

Create a deep learning classification model to predict whether applicants will be successful if funded by Alphabet Soup.

Results

Data Preprocessing

  • The target variable is the 'IS_SUCCESSFUL' column.
  • Feature variables
    Features Unique Values
    APPLICATION_TYPE 17
    AFFILIATION 6
    CLASSIFICATION 71
    USE_CASE 5
    ORGANIZATION 4
    STATUS 2
    INCOME_AMT 9
    SPECIAL_CONSIDERATIONS 2
    ASK_AMT 8747
  • The 'EIN'and 'NAME' are neither target nor features and should be removed from the input data.

Compiling, Training and Evaluating the Model

  • I have used 3 layers with 84, 69 and 41 neurons on each layer because that is what the Optimizer returned. For the activations, I used a combination of relu and tanh as recommended by the optimizer as well.
  • I couldn’t achieve the 75% target. I had tried different number of layers, neurons, and activations, but not able to get to the target.
  • Steps to increase the model performance:

    • I have eliminated the featured 'ASK_AMT'

    • I have increased the number of layers and neurons

    • I have used a combination of different activations (relu and tanh)

    • I have run the optimizer several times to find the optimal model.

Summary

  • The model can predict whether applicants will be successful if funded by Alphabet Soup with an accuracy of 72%.
  • Because of the small amount of data, a Random Forest may have a better prediction rate.
  • Eliminating more features may help improve the model.