What term is added to the weighted sum to adapt the model to data?

What term is added to the weighted sum to adapt the model to data?

• A. Bias
• B. Activation Function
• C. Weights
• D. Input

Answer: (A)

Bias is the term added to the weighted sum to adapt the model to data. In a linear combination z = sum w_i x_i + b, the bias b acts as an intercept, shifting the function so it can fit data where the relationship does not pass through the origin. This offset lets the model capture a non-zero baseline output and is learned during training along with the weights. The activation function is applied after this sum to introduce non-linearity; the bias is a separate parameter, not the activation function or the inputs themselves. If all inputs are zero, the bias sets the baseline output, illustrating its role as the intercept.

Bias is the term added to the weighted sum to adapt the model to data. In a linear combination z = sum w_i x_i + b, the bias b acts as an intercept, shifting the function so it can fit data where the relationship does not pass through the origin. This offset lets the model capture a non-zero baseline output and is learned during training along with the weights. The activation function is applied after this sum to introduce non-linearity; the bias is a separate parameter, not the activation function or the inputs themselves. If all inputs are zero, the bias sets the baseline output, illustrating its role as the intercept.