Overview

For this assignment you are going to build a report based on an algorithm you have trained. You will need to submit both a report as well as your code.

Note: You do not need to make your algorithm perfect. Use how well trained your algorithm is to provide discussion in your evaluation and conclusion.

Part $1$: Cover Page and Table of Contents

For your report, have a cover page with the title of your assignment, your name, and your class number. Have a table of contents as well.

Part $2$: Choose a Dataset.

You can either go to Kaggle or to another site that contains a set of data. You may want to choose a dataset from a topic that you like, as it will make the assignment easier and more enjoyable. Provide a link in your report to your dataset.

Part $3$: Ask a Question

In your report, ask a logistic question about your data. If you think about what we did in class, the question might be $$Would I survive the Titanic?$.$ The question you ask will be the focus of your assignment. It should be part of the title of your assignment, and what your end goal is$.$

Part $4$: Clean Your Data

You will need to go through your data removing null or empty values, removing columns that are not relevant to your question, and changing data so that the ML algorithm can process it$.$ You will need to provide an explanation for each row you are deleting or altering.

Part $5$: Graph Your Data

You need to provide at least $5$ different graphs of your data in your report. Include the graphs in both your code and your report. Make sure all graphs are properly labelled with an x and y axis as well as a title. Also add a brief description about what each graph is representing in the report.

Part $6$: Train and Test Your Algorithm

Use the data you collected to train your algorithm with a logistic regression. Make sure to split your data into a training and testing set of appropriate sizes. Include the code $($documented$)$ you used to train your data in your report.

Part $7$: Evaluate Your Model

Provide an evaluation of your model. Use what you have learned in class to decide if the model is well trained based on your test data. If your model is not within the $70\%-90\%$ accuracy range explain why you think it is not accurate. Discuss ideas of how you can improve the accuracy of your model.

Part $8$: Answer Your Question and Conclusion

Now that you have trained and tested your algorithm, try to answer your question from part $3.$ Enter data relevant to the question you asked. Use this part to also provide a conclusion to your report.