When we were looking at supervised segmentation, you worked with the following dataset: Dog No. Gender Weight Dogs? Kids? Graduated? 1 M 58 N 2 Y 2 M 73 Y 1 N 3 M 57 Y 1 Y 4 F 64 Y 2 N 5 M 75 N 0 N 6 F 62 N 0 Y 7 F 38 Y 2 N 8 M 68 N 1 Y 9 M 58 Y 1 N 10 F 62 N 1 N Now lets use some similarity and clustering concepts on this data For starters, convert the categorical variables (i.e., Gender and Dogs?) to numeri c variables by creating 0-1 variables to capture the same information. Specifically, code M as 1, F as 0, and Y as 1, N as 0. Otherwise, use the data as is to answer these questions. 1. a. Calculate the general Euclidean distance between each of these instances, resulting in an upper (or lower)triangle of distances. b. Using the hierarchical clustering process (with single linkage), iteratively join the instances together into clusters until the whole dataset is contained in a single cluster. You should be able to accomplish this using only basic Excel functions and inspection you do not need specialized software. Create a list of the clusters formed. c. Which puppy appears to be an outlier to the rest of the set? How do you know? d. Draw a rough dendrogram (it does not have to be perfectly rendered) and recommend a number of clusters (and cluster membership) for this dataset. Justify your answer. e. There is a fundamental flaw in what you have done so far. Reflect on the process, examine your answers, and make a recommendation for improving the clustering process with this data. 2. Find the 3 nearest neighbors to a new instance Female puppy weighing 50 pounds being raised in a home with 1 child and no other dogs and predict whether or not this puppy will graduate based on the majority rule AND probability estimation methods.