Imagine you have been assigned the role of chief fraud investigator for your organization. There have been several credit card fraud incidents reported. Machine Learning can be utilized to looked for unusual spending patterns and locations to help identify theft.
For this task, use this data set in the link, https://cs.coloradotech.edu/UnifiedPortal/class/293499/assignment/2657288/file/706fe772-d09b-4367-864a-c3edc98f535e/download
and create a machine learning algorithm that predicts credit card fraud. You may use any algorithm you like. Be sure to include the Python code (i.e., you must implement the algorithm in Python) to ingest, prepare, model, and show the results of the algorithm.
Submit 1 Python code file of 100–200 lines and 1 Word document.
1. You may use any algorithm you like. You must implement the algorithm in Python to ingest, prepare, model, and show the results of the algorithm. Use a training, testing, and validation approach that allows for the algorithm to be run on future unseen data.
2. Code must be well-commented. Explain your decisions and what the code is doing – it is far better to overcommit than under comment.
3. Code should adhere to best practice code standards.
4. Capture and record results and screenshots in a Word document
Also, if you would prefer to convert the file so it’s easier to follow some of the examples, see the link below.
Weka ARFF to CSV: https://pulipulichen.github.io/jieba-js/weka/arff2csv/
