ANN prediction of temperature and specific cutting force of a machining process

• This is a project based on a machining experiment performed at UNC Chharlotte, where I built different neural network architecture for a machining dataset.
• Four different architectures were considered
  • A shallow and narrow network
  • A shallow and wide network
  • A deep and wide network
  • A deep and narrow network
• ML algorithm used: Artificial Neural Network (ANN)
• The optimal ANN architecture was identified using different metrics for performance evaluation.
• Activation functions used: Rectified Linear Unit (ReLU) and hyperbolic tangent (Tanh)