Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
"Unsupervised Ensemble Learning" by Dustin Johnson #176
Unsupervised learning involves determining a hidden structure (typically groupings or clusters) from unlabelled data. Without labels, we don't have a measure of accuracy to determine which model to trust more. In this workshop, we will demonstrate how more minds work better than one and let a consensus of experts decide the ideal clustering.
Brought to you by Applied Quantitative Methods.
Time and Place
Where: Room 7010, Library Research Commons, SFU Burnaby Campus
When: Tuesday, September 12th, 2017 at 3:00 PM
I will have assumed individuals have had some experience with the implementation of an unsupervised algorithm, such as K-means, and understand the difference between the supervised and unsupervised paradigms of learning. In addition, some familiarity with statistics (averages and variances) would be helpful. We will not be covering the particular unsupervised algorithms themselves, but resources will be provided for those who are interested. The purpose of this workshop is to understand how to combine unsupervised learners together to achieve stability and robustness of performance.
Familiarity with scientific computing in Python 3 with the numpy, scikit-learn, and pandas modules would be quite helpful.
If you do not have the software above installed, please refer to the Installation section in https://github.com/jrjohansson/scientific-python-lectures/blob/master/Lecture-0-Scientific-Computing-with-Python.ipynb.
Lessons Notes: TBA