🧠 Neural Network Training Flow - Vanishing Gradient Problem

Input X:

Expected Y:

Epoch: 0

Step 1:
Random
Initialization

Step 2:
Feed Forward

Step 3:
Calculate
Loss Function

Step 4:
Calculate
Derivative

Step 5:
Backpropagate

Step 6:
Update
Weights

Step 7:
Iterate Until
Convergence

Weights/
Model

→

Input X: 1

Expected Y: 2

Predicted Y: 0.5

Loss: 2.25

Weight: 0.5

Gradient: 0.0

Learning Rate: 0.01

Problem Demonstration: Notice how the weights change by very small amounts each iteration, showing the vanishing gradient problem!