CSCI 447
Machine Learning
Spring 2019

Montana Technological University
Computer Science & Software Engineering

Schedule  |  Assignments  |  Course Syllabus  |  Moodle (447)  |  Moodle (547)

ASSIGNMENT 5

The goal of this assignment is to get experience with ensemble learning. Due date is Tues. 4/30/19, midnight. If you encounter problems, let me know. For this assignment, you can use any libraries, incuding those that implement machine learning algorithms.

1: Use different machine learning algorithms on the graduate admissions data.
In doing this, you may need to transform the data in different ways, try different machine learning techniques, and tweak hyperparameters to get the best performance. Treat this as an experiment, and document what you did in your submission file. Even attempts that are unsuccessful can be instructive/informative, so document all that you do. I would like to see at least 3 different machine learning techniques, with multiple variations of each.

2: Think about how you can combine techniques and experiment with these. Report the best results.
Again, documentation of what you think about trying and what you end up doing is critical. There are many approaches you can take here, including the random forest approach we talked about in class, or the method of building many classifiers, each in turn built upon erroneously classified examples being weighted higher. This is your chance to get creative - you want to try to get the highest accuracy possible from the data.

As you complete your assignment, please use the Jupyter notebook capability of "commenting" to document your experimaentation.

Submission: Please submit the .ipynb file(s) and any transformations of the dataset that I might need to duplicate your process to the Moodle dropbox for Assignment 5.


Page last updated: April 17, 2019