Your goal in the group project is to get the highest possible generalization accuracy on a real world application of your own creation. You will come up with some task which you believe could be generalized with machine learning and as a group you will go through all the steps from beginning to end to get a good result. A large part of the project will be deciding on and developing appropriate input features, followed by gathering sufficient labeled data. After you have come up with basic features and data, you will choose machine learning model(s), and format the data to fit the model(s). Expect that initial results may not be as good as you would like. You will then begin the iterative process of a) trying adjusted/improved data/features, and b) adjusted/different machine learning models in order to get the best possible results. You may use your own implementations of models and we encourage you to consider resources such as Scikit, Tensorflow or WEKA for doing simulations on multiple models. It is fine/great if you also want to try deep learning or other models not implemented in class for some of your models. Your written report and oral presentation (formats below) should contain at least the following:
Students taking the course for graduate credit will also have a Final Group Project, but with some additional requirements. The goal of the CS 5578 Final Project is to provide a structured opportunity for graduate students to incorporate machine learning solutions into a graduate-level research project at ISU. Machine learning is increasingly being incorporated into a variety of disciplines and the skills you gain in this class prepare you well to participate in collaborative, (potentially) cross-discipline research. The project you choose may relate to your graduate research or that of another graduate student in the course. As with the Final Group Project for the CS 4478 students, you should implement and experiment with machine learning model(s) that are appropriate to the project you choose. You are encouraged to work as a group with other CS 5578 students, but you may work alone or with a CS 4478 student (if the latter is willing to agree to the more difficult requirements for the CS 5578 Final Project).
The conference-style report for the CS 5578 Final Group project should include a related works section and compare the performance of the student's implemented model with that of other models (as implemented possibly by third-party software such as Weka) designed to solve the same task.
CS 5578 students will give an oral presentation with the same requirements as those for the CS 4478 students. Your report and oral presentation should answer the following questions:
Your oral presentation will be evaluated by your peers using this rubric. Your group should plan to take 12 minutes to present with time for Q&A to follow.