ABSTRACT
The Poisson distribution is a special type of Binomial Distribution that focuses on the average time between successes. Additionally, since the people are focusing on optimization in this chapter, it is important to understand certain topics related to differentiation with a focus toward how to apply these concept to matrices. Unlike many of the techniques presented which look to minimize a loss or cost function, the maximum likelihood technique for regression attempts to determine the interpolant with the maximum probability (or likelihood) of occurring. The gradient descent method is just one method of neural network. Neural networks are powerful tools in data science that focus on relationships in data. Activation functions are typically differentiable. This is because, in a learning process integrating a technique such as gradient descent, occurs after the activation function is applied.
