Q5 (20 marks): For the following cluster assignments, which one is better? Why? Note: Please calculate the sum of squared errors from cluster centroids to determine it. Row ID Age (years) Income (thousand riyals per month) Cluster Assignment-1 Cluster Assignment-2 A 0.2 1 C1 C2 B 1 0.6 C2 C1 C 0.2 0.8 C1 C2 D 0.3 0.7 C2 C2 E 0.9 0.8 C1 C1 Sum of Squared Error for Assignment-1: Sum of Squared Error for Assignment-2:
Added by Marina F.
Close
Step 1
C1: A(0.2, 1), C(0.2, 0.8), E(0.9, 0.8) Centroid_C1_x = (0.2 + 0.2 + 0.9) / 3 = 1.3 / 3 = 0.433 Centroid_C1_y = (1 + 0.8 + 0.8) / 3 = 2.6 / 3 = 0.867 C2: B(1, 0.6), D(0.3, 0.7) Centroid_C2_x = (1 + 0.3) / 2 = 1.3 / 2 = 0.65 Centroid_C2_y = (0.6 + 0.7) / 2 = 1.3 / Show more…
Show all steps
Your feedback will help us improve your experience
Juhi Singh and 72 other Intro Stats / AP Statistics educators are ready to help you.
Ask a new question
Labs
Want to see this concept in action?
Explore this concept interactively to see how it behaves as you change inputs.
Key Concepts
Recommended Videos
a) Given are the two dimensional points A(2, 6), B(3, 5), C(4, 7), D(9, 2), E(5,10), F(11,7). Compute single-link bottom-up hierarchical clustering using the Euclidean Distance formula. Show the clusters representation in xy-plane and draw a dendrogram of it. b) Given are the points A = (2,3), B = (3,3), C = (3,2), D = (-2,5), E = (-3,-2), F = (-2,-2). Starting from initial clusters Cluster1 = {A} which contains only the point A and Cluster2 = {D} which contains only the point D, run the K-means clustering algorithm and report the final clusters. Use L1 distance as the distance between points which is given by d ((x1, y1), (x2, y2)) = | x1 – x2 | + | y1 – y2 |. Draw the points on a 2-D grid and check if the clusters make sense. You may stop the clustering process if it found that two iterations have the same clusters, otherwise, at least 4 iterations are required.
Juhi S.
Q5) [10 marks] Consider the following data set D={ (2,0), (2,1), (1,2), (3,2), (2,3), (3,3), (2,4), (3,4), (4,4), (3,5) } (a) Apply K-means clustering with k=2, using Manhattan distance. Start with (2,0) and (3,5) as the initial centroids (seeds). Show all intermediate clusters in a table, as well as the final clusters. (b) Calculate the Dunn index at each step.
Adi S.
(1) Consider the data shown below. There are two groups, depicted as open dots (group Y = 1) and solid dots (group Y = 2). (a) Suppose that misclassification costs are equal $c_1 = c_2 = 1$. Which of the following clustering schemes is likely to perform the best? • Fisher's linear discriminant analysis; • K-nearest neighbor with K = 5; • K-nearest neighbor with K = 50. For the other two schemes, explain why they are not likely to perform very well. (b) Both sampling distributions $X|Y = j$, for $j = 1, 2$, appear to be approximately rotationally symmetric. (Recall that Y is rotationally symmetric if $z^T Y$ has the same distribution for all $z \in \mathbb{R}^2$ with $|z| = 1$.) Suppose that, for this sample, we have that $(\bar{x}_1 - \bar{x}_2)^T S_{pool}^{-1} (\bar{x}_1 + \bar{x}_2) \approx 0$. Find the approximate expected misclassification cost if we classify data using Fisher's linear discriminant analysis. (c) Suppose we instead classify based on $|X|^2$. Assume that $\pi_1 = 0.45, \pi_2 = 0.55$, and that $|X|^2 | Y= 1 \sim Exp(1)$ and $|X|^2 | Y=2 \sim Unif[4, 5]$. Find the expected misclassification cost under the optimal classifier, amongst all rules based on $|X|^2$ . (d) Explain why it can be useful to first implement an agglomerative clustering algorithm, before performing a K-nearest neighbor classification scheme.
Lainey R.
Recommended Textbooks
Elementary Statistics a Step by Step Approach
The Practice of Statistics for AP
Introductory Statistics
Transcript
Watch the video solution with this free unlock.
EMAIL
PASSWORD