Question

Q4 (4 marks) (a) What is machine learning? (b) Discuss the difference between classification and clustering. Give one example algorithm (6 marks) for classification and clustering respectively For Q4(c)-Q4(f, consider the figure below which shows the examples (instances) with different Yellow Red Purple Orange Height Blue Red Violet Green Width (c) Is k-Nearest Neighbors (KNN) a classification method or clustering method? mark) (d) What is the outcome of KNN for the query point based on I-nearest neighbor? mark) (e) What is the outcome of KNN for the query point based on 2-nearest neighbors? mark) () What is the outcome of KNN for the query point based on 5-nearest neighbors? (1 mark) For Q4(g)-Q4(h), consider the following transaction table. TID I Items 1Bread, Milk, Chips, Mustard Beer, Diaper, Bread, Egg Beer, Coke, Diaper, Milk Beer, Bread, Diaper, Milk, Chip 5 Coke, Bread, Diaper, Milk Beer, Bread, Diaper, Milk, Mustard Milk 7 Coke, Bread, Di -> (Diaper If both the (g) Find the support and confidence for the association rule (Milk) thresholds for minimum support and minimum confidence are 60%, does this rule qualify? Why? If both the thresholds for minimum support and minimum confidence are 60%, does this rule qualify? Why? (3 marks) (h) Find the support and confidence for the association rule (Coke, BreadDiaper, Milk ) (3 marks)

3112 1617 Q4

Answer 1

4 a)machine learning is a technique which enables the computer to predict outcome from the previous experience without being explicitly programmed

b)classification is a kind of machine learning algorithms in which data is classified into previously defined classes whereas in clustering algorithms there are no predefined class labels

clustering k-means is used to cluster the unlabelled data into k groups.The algorithm iteratively assigns data point to one of the k-groups based on the distance between the data point and the centroids of cluster

classification knn is classification algorithm in which when a new data point comes .The algorithm find k nearest data point to it.and assign it the label which majority of these k-data points are holding

c)knn is a classification algorithm

d)if the value of k is 1 it will find the nearest point to U and assign it the same label.As the nearest point to U is holding label orange the U will be labelled as orange

e)the two nearest neighbours of U are red and orange as there is a tie U can be assigned either of them .To deal with such situation k is usually odd

f) The 5 nearest neighbours of U are orange,red,red,violet,blue and among these majority is red so U will be classified as red.

As the HomeworkLib policy asks to answer first four parts of the question with multiple parts .but i have answered 6 question

If you have any doubt with these connect in comment section.Please do upvote.