Describe your clusters in english

Assignment Help Computer Engineering

Reference no: EM133324775

Assignment: In this assignment you will perform some basic clustering on your dataset.

First, use k-means clustering. Ignore your class variable when constructing the clusters. Evaluate the results on the training set. Vary the number of clusters k, and choose the one that you believe gives the "best" results. Unless you have strong, clear domain knowledge, you should justify your choice of k with computational results.
Describe your clusters in English. Your description will be based on the cluster centers, but use descriptive terms instead of just listing the center values. Be concise.
Use the cluster visualization feature in Weka to find one or two two-dimensional projections of the data that show the clusters clearly. Note that you may need to turn on the jitter in order to clearly show discrete features.
Now use the class variable to evaluate the clustering (using the value of k chosen above). How well did the clusters classify? Try a higher value for k - does it work better? What can you conclude from this? (Note: do *not* cluster using the class variable - use it only for evaluation.)
Finally, repeat using EM clustering. This time the software will choose the value of k. Does it agree with the one you chose earlier? How well did EM classify (that is, predict the class variable) as compared to k-means? Describe the clusters as above, this time taking into account the standard deviations. Which explanation do you find to be the best?

Reference no: EM133324775

Questions Cloud

What could be the source of infection : Create your own any fictitious security attack scenario at any organization that requires an immediate attention. Use a security awareness training as the first

How would the effectiveness of the actions : How would the effectiveness of the actions/improvement initiative would be evaluated. In other words, how would you know you were successful

What would have been your response as a nurse to mrs charles : What would have been your response as a nurse to Mrs. Charles when her husband was actively dying and she felt overwhelm and wanted him fed

Write a program in mps430 assembly language subroutine : write a program in MPS430 assembly language subroutine that examines the contents of register R11 and looks for a 3-bit sequence "101". The program will count

Describe your clusters in english : Describe your clusters in English. Your description will be based on the cluster centers, but use descriptive terms instead of just listing the center values.

Describe a scientific misconduct : DNP 810 Aspen University Describe a scientific misconduct within the last five years involving some aspect of the process of conducting research

Identify nft economy for sustainability of music production : Identify an advantage and a limitation of the NFT economy for the sustainability of music production in today's web3 environment.

Is this the inclusion or exclusion criteria : significant functional or severe cognitive impairment, or both, and (iv) the prioritization of symptom control above disease progression

Write function that takes two arrays : Write function that takes two arrays, corresponding to two subsets of numbers, and returns true if each array has the same sum, and returns false otherwise.

User Account

All Pages