QUESTION 5 Which statement about k-NN is FALSE? A. If k in KNN is too high, we might miss out on the method's ability to capture local structure, one of its main advantages. B. If k is too low in KNN, we may be fitting to the noise in the data. C. Higher values of k in k-NN provide smoothing that reduces the risk of ovrefitting due to noise in the training data. D. Because the parametric assumptions underlying k-NN are many, a lot of time is required to estimate parameters from the training data. 6 points

QUESTION 6 Which of the following is FALSE about using categorical variables in k-NN? A. Before k-NN is applied, do not convert the variable to binary dummy format. B. Convert categorical variables to binary dummies before applying k-NN. C. k-NN does not come with multicollinearity problems that can be found in multiple linear regression models. D. The euclidean distance is a popular measure used to compute distance with k-NN classification for categorical variables converted to binary dummies. 6 points

QUESTION 7 Predictors are reduced by using methods such as singular value decomposition, principal components analysis, and factor analysis. However, if the number of dimensions grow, a phenomenon occurs known as the curse of dimensionality. What is the curse of dimensionality? A. It means the records are grouped together based on their distance. B. It refers to the fact that in k-NN, the distance is computed from the dataset during the time of prediction. C. It refers to the fact that the distance no longer has to be calculated when dimensions grow in k-NN. D. As the number of predictors p grows, the number of records that we need to use k-NN accurately grows exponentially.

6 points QUESTION 8 The logistic model involves multiple steps, the first being understanding probability (variable p). The probability of winning something and odds of winning are closely related. Odds are used for predicting horse races or gambling outcomes, If the odds of winning such an event is 4:1, how can you manipulate the formula below to find p? Odds(Y=1)=P/1-P A. .125 B. .60 C. .80 D. 20