Data Mining Project Description

Data Mining Project Description

CIS 4321 Summer 2020

In this project, you experience the cycle of the data mining predictive process. Below, I explain the different stages of the project.

Requirements

For the final project in CIS4321 , you are going to mine a dataset and define a project scope, implementation, and analysis.

1. The focus on the project should be on predictive modeling in the business domain using either regression, classification, or clustering.
2. The dataset should be interesting, non-trivial and should have at least 6 attributes and on the order of 1000s (or more) instances.
3. The data mining process should include data preprocessing (as needed), exploratory and descriptive analysis, visualization and model building and evaluation.

THE DELIVERABLES

You will be responsible for submitting the below report in the specified outline and format.

Project Report:

The introduction is where you provide a brief description of your personal motivation for the project and frame your report. Tell the reader why they care about the results you are about to present and what is the question you will be answering is important. A description of your dataset including what type of data it contains, how many attributes, how many instances, what the class labels are if its classification. Any additional challenges such as messy or missing values.

The data analysis section is where you describe your data showing exploratory analysis through charts and tables showing basic data distributions (e.g., frequency diagrams, scatter plot diagrams), and “interesting” relationships between attributes. Your job for this part is to give more insight into your dataset that is relevant for the predictive modeling task. Explain what the data is and anything about that data that is especially interesting.

Modeling tasks, this section should describe the modeling approach you developed and its performance. Explain what techniques you used, briefly how you designed and implemented the model, how you tested the predictive ability, and how well it performs. Explain the method(s) you used in the data mining process. Discuss how the method works, why it was well suited for your data, and how you applied it.

The results section is where you describe what the analysis resulted in. Again, it is generally appropriate to use charts or tables here to illustrate important findings.

The discussion section is the place to explain your findings. Why do you think you found the results you did and what do you think they mean? This is generally the only section where it is appropriate to make educated guesses about the phenomena you observed. A brief description of your new data mining topic you learned.

Finally, the conclusion should provide the analysis’ answer to the question posed in the introduction along with a brief description of why the analysis answered the way it did, which should be consistent with your discussion section. Additionally, you may wish to posit questions raised by your analysis or areas for future analysis. In data analysis reports, it is also common to have a recommendation. Describe any key components of the code.

The document should be in APA format, between 5-8 pages in length  (12 pt font, double-spaced), not including title page, content page, or images/graphics.

Project Submission

Submit your project to blackboard by the due date, no late submissions will be accepted.

You should submit:

1. Your source code in a well-documented Jupyter Notebook and dataset files.
2. Your project report First_Lastname_FinalProject.docx

Grading Guidelines

This assignment is worth 100 points. Your assignment will be evaluated based on a successful compilation and adherence to the program requirements. I will grade according to the following criteria:

• 50 pts for implementation
  • 25 pts for relevance to the methods learned in the class /originality of project
  • 25 pts for technical rigor and complexity
• 50 pts for technical writing, reporting and format.