Little Bits of Artificial Intelligence

To really understand Gradient Descent, we should know:

1. Linear Equation Basics:

• Understand y = mx + c (we use y = wx + b in neural networks)

2. Mean Squared Error (Loss Function):

• MSE = average of (predicted – actual)²
• Helps us know how “bad” the prediction is.

3. Derivatives / Slopes:

• How a small change in weight affects the loss
• We don’t need to do calculus, but knowing that derivative = slope helps.

4. Minimization:

• We want to reduce the loss → we follow the negative slope

5. Learning Rate:

• A small number that controls how fast or slow you move in the right direction.

Summary Visual (Text Style)