CSCI 447 |
1: Run the Machine Learning Walkthrough - Use the provided banking data
The slides posted for the lecture on 1/18/19 show a walkthrough of this process. You can use those to follow along.
2: Run the Machine Learning Walkthrough - Use graduate admissions data
Once again, you can use the slides from lecture to follow along, but you will need to upload your own data and choose appropriate settings (identifier column, name row, output variable) and then run the ML model as before. Don't expect to get the same good results that you got on the banking data. See, there really is a need for data scientists - the automated approach doesn't quite work...
3: Do the Jupyter Notebook Tutorial - Use graduate admissions data from the last part
You may either create a notebook on AWS or download something like Anaconda on your own machine and do it locally. If you do it locally, please submit your notebook to the Moodle dropbox. I am hoping enough people do it on AWS that I can test how that works from the instructor point of view. The link to the notebook tutorial is in the lecture slides.
4: Go to SageMaker, create a notebook instance (if you haven't already), and use Jupyter to browse through existing notebooks
There is nothing to turn in here, just look at some of the sample notebooks in the SageMaker Examples tab so you get an idea how the machine learning process is conducted using SageMaker.
Page last updated: January 18, 2019