[ML Interview] - What’s the trade-off between bias and variance?

What’s the trade-off between bias and variance?

• Bias and variance are complements of each other.
• Model with the higher bias tends to under fit the point of the dataset but has smoother inference curve.
• Model with the higher variance tends to over fit the point of the training data but takes into more consideration the noise and fluctuation.
• When considering the bias and variance trade-off, we want to find the optimal point which has both low bias and low variance by tuning the model complexity.