1.Build a linear model to forecast a number of total rentals ( count ) using potential predictors,...
Question:
1.Build a linear model to forecast a number of total rentals (count) using potential predictors,season, holiday, workingday, weather, atemp,andregistered. Note here the linear model is not ideal for predicting count. We canwork around this drawback by rounding up or rounding down the predictions
Perform best subset selection using bestglm() function based on BIC. What's the best model based on BIC?
3. Compute the test error of the best model based on BIC using LOOCV.
4. Calculate the test error of the best model based on BIC using 10-fold CV.
5. Perform best subset selection using bestglm() function based on CV. What's the best model based on CV?
6. Perform the backward stepwise selection using stepAIC() function. What's the best model?
dataset- https://github.com/MyGitHub2120/bike