Wed, 03 Jun 2015 06:55:39 -0700
When you think of interns, you might first conjure up the image of a frazzled student fetching coffee and scheduling meetings. Maybe it’s because we already know how to work the Keurig ourselves, but we don’t subject Chitika interns to “gofer-ing.”
Instead, Worcester Polytechnic Institute students Youwei Hu and Jeremy Macaluso focused on things like real-world applications of binomial equations and order statistics for the past six months, as part of their WPI Major Qualifying Project.
Working closely with Chitika Data Engineer Joe Regan, Hu and Macaluso worked to improve Chitika’s self-serve advertising platform, Cidewalk. Their goal: Optimize Cidewalk’s automated bidding process to win the most impressions possible, at the most profitable rate.
In their endeavors, Hu and Macaluso went through many possible models to find the general optimal bid.
It wasn’t an easy task: some of the most frustrating moments included attempts to simplify a formula that was just too large to deal with. Instead of continuing the headache, they decided to go a different route that eventually lead them to their ideal analytical formula.
On Wednesday, the students gave a presentation on their research and conclusions, finally ending with a suggestion to fully integrate the formula they came up with, after a few minor tweaks.
Moving forward, the Cidewalk team will be working with Hu and Macaluso to integrate their solution into the final product.
Hats off to Youwei and Jeremy for sticking with the project and helping us find a real-world solution to a real-world problem!
Instead, Worcester Polytechnic Institute students Youwei Hu and Jeremy Macaluso focused on things like real-world applications of binomial equations and order statistics for the past six months, as part of their WPI Major Qualifying Project.
Working closely with Chitika Data Engineer Joe Regan, Hu and Macaluso worked to improve Chitika’s self-serve advertising platform, Cidewalk. Their goal: Optimize Cidewalk’s automated bidding process to win the most impressions possible, at the most profitable rate.
In their endeavors, Hu and Macaluso went through many possible models to find the general optimal bid.
It wasn’t an easy task: some of the most frustrating moments included attempts to simplify a formula that was just too large to deal with. Instead of continuing the headache, they decided to go a different route that eventually lead them to their ideal analytical formula.
On Wednesday, the students gave a presentation on their research and conclusions, finally ending with a suggestion to fully integrate the formula they came up with, after a few minor tweaks.
Moving forward, the Cidewalk team will be working with Hu and Macaluso to integrate their solution into the final product.
Hats off to Youwei and Jeremy for sticking with the project and helping us find a real-world solution to a real-world problem!
0 Comments