Neural_Network_Charity_Analysis
Data Analysis and Visualization 2022
Charity Spending: Determining Likelihood of Success with Machine Learning.
Results:
Data Preprocessing
- The features of this model include data relating to the type of application recieved.
- Some data, such as identification factors, were not needed and could be dropped.
- Model targets include the factor of weather or not the funding was successful.
Compiling, Training, and Evaluating the Model
- For the optimized model I chose a model known as the feedforward network, having 16 layers, alternating normalization, dense, and dropout layers with decreasing complexity.
- The model achieved over 60% accuracy but did not attain the desired outcome.
Summary: The results of the machine learning model indicate over 60% accuracy in predicting successful outcomes.