Data Scientist Preparation

1. The trade off between Bias and Variance.

To build a model, in general, there are two types of errors, the first one the basic fundamental assumption error. The second one, the unseen data point fitting error.

High bias, in general mean the system try to simplify the problem by adopting a simple model to fit the observed data, which will make the model less variant to the future data points.

High Variance model, in general has low bias. The model tries to capture all the detailed signals. And this tends to overfit the current data. And hence the variance is higher.

2.