Please provide the inline answers for the following 3 questions:
Question 1. I’d like to understand why you have not taken loss function into the account while resolving the challenge. ?
Questions 2 Is Loss only used for the informational purposes in the earlier examples as well i.e., in Gradient Descent: The Batch Update and Gradient Descent: The Stochastic Update?
Question 3. Directly calculating the gradient without calculating the loss would serve the purpose of updating the weight also serves the purpose of perceptron training in the previous two examples as well?
Hi @Vikrant !!
1: The perceptron algorithm used in the code does not involve a loss function as it focuses on updating weights based on the correctness of predictions.
2: In the earlier examples, the loss function was used for informational purposes and to quantify the model’s performance during gradient descent optimization.
3: The perceptron algorithm directly calculates the gradient without explicitly calculating the loss, serving the purpose of updating weights, but lacks a measure of model performance like a loss function.
I hope it helps. Happy Learning 