Exponential Distribution and Gradient Descent
The Exponential Distribution and Gradient Descent notebook models the elapsed time between events, such as the time between each ad clicks or video views with an exponential distribution. For any given , the distribution gives the likelihood that time passes between two consecutive events. The notebook uses tensors to hold the data, and the optimization algorithm to minimize the problem is a gradient descent algorithm, which uses DiffKT for automatic differentiation.