Your Task: Our objective is to construct classifiers that can precisely identify hospitals with ratings below the U.S. average. After thoroughly examining the data and eliminating missing values as "Not Available," please answer the following questions: Part I: Data Exploration (Notes: read through the data dictionary in Table 1) a) [10 points] Which type of hospital ("Acute Care Hospitals", "Critical Access Hospitals", or "Acute Care - Veterans Administration") is more likely to have a low rating (on average)? Please show evidence. b) [10 points] Do hospitals with Emergency Services have a better rating than the others on average? Please show evidence. c) [10 points] Imagine you received two proposals for building a new hospital. All things equal, if one of them proposes a "Government-State" and another proposes "Government- Local", which one would you approve of? Why? d) [10 points] What are the best and the worst states in terms of the proportion of hospitals with low quality ratings? Please mention the five best states and five worst states e) [15 points] Build a decision tree model and find the minimum error tree. Please report the accuracy, confusion matrix, and roc curve of the model. Predicted 0 Predicted 1 f) [15 points] Build a random forest model and set the number of estimators as 50. Please fine-tune the hyperparameter max features in the random forest model. Which value of max features would you recommend? Please report the accuracy. confusion matrix, and roc curve of the model. 0 1 Reference 0 Reference 0 1 ) (10 points] Based on your analysis, which model do you recommend? Can we achieve a high level of predictive performance in predicting low-quality hospitals in the U.S. using this data set? Then discuss if the data is good for this task or not. h) (10 points) Based on the results from the optimal model, identify the most