Using resampling approach of your choice (e.g. cross-validation, bootstrap, etc.) obtain test error as well as sensitivity and specificity for each of these methods (logistic regression, LDA, QDA, KNN with $k=1,7,55,351$).Present results in the form of boxplots, compare test error/sensitivity/specificity across these methods and discuss their relative performance.

**Important note:** : dataset: Wifi_localization (https://archive.ics.uci.edu/ml/datasets/Wireless+Indoor+Localization)

we will be predicting whether the phone is at location=3 or not, as opposed to working with multi-class predictor. In other words, before you proceed with any of the problems in this homework, please convert the four-levels outcome to the outcome with only two levels: location=3 (must be 500 of those) and not (must be 1500 of them).*

*If you are creating a new column containing this binary outcome, please make sure that the original outcome with four columns is NOT used inadvertently as one of the predictors. If you are getting invariably 100% accuracy regardless of the choice of the method or split of the data into training and test, chances are your code is using original four-levels outcome as a predictor.*