K-Means Clustering
Part 1 -
1. Introduction - Cluster analysis involves grouping things together so that the members of each group are more similar to each other than to members of other groups. There are numerous algorithms or models associated with clustering such as k-means clustering, hierarchical clustering, and density models.
Cluster analysis is popular in market segmentation. For example, the market for a product or service may be segmented into groups of customers or regions that share common interests or are similar in terms of their preferences and socio-economic attributes. An appropriate marketing strategy may then be devised to serve the needs of identified segments better.
This part involves using k-means clustering as a clustering tool to be applied to a data mining study within your domain of interest using R and RStudio.
2. Steps to Completion - For each study the general procedure is to:
- Review theoretical background based on available resources in the course content
- Select a dataset from the module's recommended datasets list
- Run an analysis, perform evaluation, and capture the results
- Document your findings and analysis in a data mining analytical report
3. Deliverables - Submit your analysis report by addressing the following critical areas:
Introduction: give some background and context about the domain of application, provide the rationale for the type of analysis, and state the objective clearly.
Analysis: describe the data both qualitatively and quantitatively through exploratory analysis, perform necessary preprocessing activities, give some intuition about the algorithm and core parameters, demonstrate the model building steps along with parameter tuning, and explain all your assumptions.
Result: explain the result and interpret the model output using terms that reflect the application area, perform model evaluation using the appropriate metrics, and leverage visualization.
Conclusion: summarize your main findings, discuss experimental limitations related to the data and/or implementation of the algorithm, and suggest improvement areas as a potentiation future work.
Miscellaneous:
- Proof read your report for correct structure, grammar, and spelling
- Follow appropriate APA formatting and provide all references
- Include your R script and extended model outputs in an Appendix section.
The length of the report should be 7-10 pages excluding the title page, appendix and R script.
Part 2 -
Run an exercise on a vehicle dataset and write a report on your findings and results interpretation in your own words. The report needs to cover the exercise key points below in order.
Download the vehicle.csv file to your hard drive.
1. Introduction - What do you expect the k-means clustering method to accomplish for the vehicle data?
2. Data pre-processing
- Run the set.seed command. Include the command on the report and explain the reason for running this command.
- Load the data from vehicle.csv file into R. Create a copy of the vehicle dataset called myvehicle. Include the command in the report.
- Remove the variable class from a myvehicle. Include the command in the report, and explain why we remove the class variable.
- Run the scale command to scale the myvehicle. Include the command in the report, and explain why we scale data.
- Discuss any additional data pre-processing that you run. Include the commands and explain what each command does in the report.
3. Run the kmeans method with k=4 and store the output in the variable kc. Include the command in the report and discuss the input parameters you used. Enter kc at the command prompt and hit enter. Include the command output in the report and answer the following questions.
- How many instances are in each cluster?
- What information does the cluster means section of an output provides and how were the numbers obtained?
- What is clustering vector?
- What is sum of squares by cluster, and what does it mean?
- Run the kc$iter command, and explain what the output shows. Include the command, the output, and explanation in the report.
4. Clustering evaluation
Build the cross-tabulation to compare how the method clustered the vehicles with the actual vehicle class. Include the command and the output in the report. Answer the following questions.
- What is the dominant vehicle class in each cluster?
- What additional information does the table show?
- What percentage of vehicles were clustered in agreement with the actual class?
5. Build the cluster plot. Include the command, the plot, and the plot interpretation in the report.
6. Experiment with 3 different k values, and summarize the findings in the tabular format.
|
k
|
Number of instances in each cluster
|
Between clusters sum of squares
|
Within clusters sum of squares
|
Number of iterations
|
|
4
|
|
|
|
|
|
Value of your choice
|
|
|
|
|
|
Value of your choice
|
|
|
|
|
|
Value of your choice
|
|
|
|
|
Explain the effect of k values on method results.
What is an ideal value of k for the vehicle data? (This is an open-ended question)
7. Summary
- What differences between k-means clustering and classification methods did you observe?
- Which part of this exercise did you find the most challenging and which approach did you take to resolve the challenge?
Attachment:- Assignment Files.rar