Dylan Sosinsky Week 3– Model Development
Dylan S -
Hey blog,
After successfully converting the multi-linear regression from R to Python for the 10 days out model, I repeated the same process for the other models (14 days, 7 days, 5 days , 2 days, 1 day). The coefficients matched the ones on the spreadsheet, so it was a big success! Also this week, I decided to explore the “new features” that the team had started adding to the model like “temp.num(temperatures categorized in ranges),” “fall break,” and “holiday break.” I successfully added these new features to my models in python and once again matched the coefficient values on the spreadsheet.
One of the nice things about the spreadsheet we use is that we can put built in equations for certain values were trying to calculate based on values in the spreadsheet cells. In our case, since we know a linear regression is: constant + coefficient(x)(value of feature (x) +….., we can actually implement this into our own spreadsheet. I fear that the equation in the spreadsheet may be a little off though as it seems to be generating different predicted attendance than the models and hand written equations that were thoroughly reviewed. Thus I will have to try to find the error if there is one with the equation in the spreadsheet.
That is all for the week.
Thank you.
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.