This consists of the following works:
- Creating a classification model using a Logistic Regression with a Neural Network mindset
- A sample neural net with a 1 hidde layer that is used for a image classification problem
- A deep neural net with multiple hidden layers that has better accuracy when compared to above problems
Concepts covered are:
- Initialization of the Weights and biases
- Forward Propagation: Linear and Activation using different activation functions
- Backward Propagation: derivatives using the cache
- Compute Cost: uses the loss function to come up with the cost
- Updating the parameters: leverages the learning rate too
- Predict using the model
- Cost Optimization Analysis