First of all, in terms of a report, I really like it that the report looks very professional, for example, it has very good format, abstract, and figure names. It gives readers a clear view of what each section is about.
Secondly, I like the group are listing two models they use. The random forest and logistic regression. And I think the model and the explanation look reasonable.
I like the future step the group is talking about. They are using some of the knowledge outside of the class, including LOOCV and SMOTE, which I think is very good. I agree that it would be a good idea to add in the SVM datasets.
What I think the group could do to improve in the report
First of all, the group introduces what types of data are their feature variables, but I would suggest the group can give a brief description of feature variables that they keep after data cleaning, in order for readers to have a better understanding of the report.
In random forest example, the group could explain more about how they choose the tree and why they restrict the depth to be 30 out of 55 features.
I would suggest the group to talk about more in their data cleaning step, such as whether they get rid of empty data, how they transform Boolean, string variables to integers etc.
There are three things I like about this project.
What I think the group could do to improve in the report