Task2: Linear regression: This is a dataset to predict the heath insurance costs of customers for a real health isurance company For "sex", and "smoker" features, replace classes with 0 or 1. For instance "female" could be 0, "male" could be 1. Choice of 0 Use multiple linear regressionto predict charges. Set 20% of the data to be "test" set. Calculate RMS for test set and training Find out which feature can be dropped, either using correlaton between features, or use t-stat ot p-value. Build a new model Compare first model and second model. Which one is better? Why? You can create as many sheets aas you want for this task, name them: Task2-1, Task2-2, ... . I already have created two blank s