Discrete vs. Continuous Prediction
Explore the concepts of discrete versus continuous predictions in deep learning models. Understand how step functions provide binary outputs while sigmoid functions offer probabilistic outputs. Learn to apply these activation functions to perceptron models to classify data effectively.
We'll cover the following...
We'll cover the following...
Discrete vs. continuous predictions
The prediction is the answer we receive from the algorithm. A discrete answer would result in the output of “yes” or “no”. And, a continuous output would be a number between 0 and 1.
Consider the classification problem where it is required to draw a decision boundary that separates the stars from squares. A discrete algorithm would tell us whether the class is stars by predicting 1 for stars and 0 for squares ...