Here is my Research Topic: Predicting Bank Failures in USA Here is my Research

Here is my Research Topic:
Predicting Bank Failures in USA
Here is my Research Question:
Which U.S. state is most likely to experience the greatest number of bank failures in the future, and what are the main factors that contribute to this likelihood?
Here is my Dataset
Dataset Name: FDIC Failed Bank List
FDIC Failed Bank List Dataset
Download the FDIC Failed Bank List (CSV)
Here is my chosen ML Method:
Classification machine learning method
Revise and Evaluate Data Analysis Model
In this milestone, you will perform an evaluation of your data analytic model and revise your decision model as needed. For the revision, you can add in additional machine learning models or do feature engineering. You can also create confusion matrices and check for accuracy, precision, recall, or F-Measures. You can do sensitivity analyses, create ROC curves, check error rates and variable selection/feature selection. Please do see the image below for some other options for revisions:
Deliverable
For milestone 4, please ensure you have the following REQUIRED sections ONLY:
Final Research Question: Please state your final research question and describe how it evolved, if it changed, from your Milestone 1 version.
Your final research question should reflect the actual analysis you conducted and the actual insight or prediction you made. For example, if you actually ended up doing a classification task to predict the value of a response variable based on some other predictor variables, that’s what you should state as your precise, quantitative research question.
Model Revision: You should discuss how you revised your model(s) and perhaps how you narrowed down the scope of the project to something coherent and managable. In addition, you should give the details of your final machine learning model and present your final results.
Concomitantly, you should assess the robustness of your model, especially noting what makes it strong or what breaks it or the distribution of uncertainty in it. As such, you’ll likely want to utilize some variant of sensitivity analysis, depending on your particular model, to demonstrate true understanding of what the model’s doing rather than just a rote implementation thereof.
Final Results and Initial Version of Final Conclusion: You should show the initial validation of metrics and also decide upon any additional machine learning algorithms you might want to try. In addition, please ensure you answer the following questions:
What did you learn about the data?
How did you answer the questions?
How can you justify your answers?