Question 4 Which of the following statements are true about Principal Component Analysis (PCA)? (a) PCA works by finding lower dimension embedding that matches those pair-wise distance of the original representation. (b) PCA works by finding the direction of projection so that the lower dimensions data has maxi- mum variance (c) PCA is much slower to run than t-SNE (d) PCA uses a Non-Linear projection Question 5 Which of the following statements are true about t-distributed stochastic neighbor embed- ding (t-SNE)? (a) t-SNE works by finding lower dimension embedding that matches those pair-wise distance of the original representation (b) t-SNE works by finding the direction of projection so that the lower dimensions data has maximum variance (c) t-SNE is much faster to run than PCA (d) t-SNE uses a Linear projection Question 6 Which of the following statements best describes the generative model in the context of clas- sification? (a) In Generative model, one can build distribution for additional class without re-train data of existing classes (b) Let x, Y and be the data, label and parameter respectively. Then in Generative model, one may compute p(Y|x) directly without the knowledge of p(x0). (c) Generative model needs to re-train all classes when a new class data is introduced. (d) Generative model usually works by directly distinguish boundaries between classes Question 7 In a joint step of apriori algorithm, let A = {a1, a2, ak-1} and B = {b1,b2, bk-1} be a pair of frequent (k-1)-itemsets from L(k-1). A and B are merged to Ck if they satisfy which of the following conditions: (a) a = bi( for i=1,2,..., k-2) and ak-1 bk-1 (b) a = b (for i=1,2,..., k-2) and ak-1 = bk-1 (c) ai bi( for i=1,2,..., k-2) and ak-1 = bk-1 (d) abi (for i=1,2,..., k-2) and ak-1 bk-1 Question 8 for the following transactions Transaction ID 2 3 4 Items Bought A, B, C A, C A, C B, E, F which of the following is true about the confidence for AC and C A: (a) A C is 100%. (b) A C is 66%, (c) AC is 100%, (d) A C is 33%, CA is 100%, CA is 100%, CA is 66%, CA is 100%, 2 2/4