Industrial Projects

2014-2015 Projects

JPMorgan Web Service Registry

Sponsor: JPMorgan
Advisors: Jon Abraham, Kevin Sweeny (MG), Arthur Gerstenfeld (MG), Micha Hofri (CS), Xinming Huang (EE)
Students: Heather Jones, John Bosworth, Ibraeim Bukhamsin, Arin Kulvanit, Jordan Wetzel

To assist JPMorgan employees in searching for a company created service application, our team created an internal service registry from scratch to organize existing services in one central location. Our team developed a user-friendly front-end web-based application that displays current services as well as allows employees to add new services. We also focused on designing a back-end for interfacing with our newly designed database that stores all information about current services.

Liability Driven Investing – Investment Process

Sponsor: JPMorgan
Advisors: Jon Abraham, Kevin Sweeny (MG), Arthur Gerstenfeld (MG), Xinming Huang (EE)
Students: Qiao Li, Tianhai Chang, Saad Riaz

Companies need assistance with their retirement plans in order for them to pay off their liabilities. JP Morgan Chase & Co. is a corporation that encompasses a wide range of financial aspects. Retirement plans is a crucial one. It is vital for JPMC to be able to make such monetary facets easily understandable for its clients. Our group worked on the investment process that is driven by the liability costs of JPMC’s clients. Each one of us worked in a different area; The Funds of Funds is where the client's portfolios are created, in the Liability Driven Investment the discounted liability and asset is calculated, and the Data Warehouse is what provides the necessary data.

A Comparison of OLAP and Coherence Large Scale Aggregations

Sponsor: BNP Paribas
Advisors: Jon Abraham, Kevin Sweeny (MG), Arthur Gerstenfeld (MG), Xinming Huang (EE)
Students: Kimberley Tate, Jiahui Li, Lili Zhang, Xinyue Zhong

The goal of the project is to develop a generic mechanism for cache-based aggregation of data. This will allow the end user to have a summary view with multiple pivot points. To achieve this goal, the following steps will be taken: 1. Converting the database to files 2. Loading the files into caches 3. Aggregating the data in the caches Coherence and aggregators will be used to complete these tasks. To write the code for these processes, Java, C# and SQL are the programming languages that will be used.

Pre-Mission Flight Plan Optimization

Sponsor: Draper Laboratory
Advisors: Andrew Trapp (MG), Suzanne Weekes
Students: Ethan Moon, Alexander Sunde-Brown

The goal of this project is develop a tool to help with mission planning undertaken at the National Aeronautic and Space Administration in partnership with Draper Laboratory. The aim of these missions is to gather data using Earth- observing aircraft at a set of sites. In this project, we aim to minimize the total mission time. We model this minimization problem as a variant of the Distance Constrained Vehicle Routing Problem. We present two integer programming formulations of the problem as well as a supporting proof. While the results we present show that the solution time is limited by the available flight time for each aircraft and the number of vehicles, our implementation of the integer programming problem is able to solve problems with up to 35 sites.

Optimal Bid Pacing for Online Ad Impression Vickrey Auction markets

Sponsor: Chitika Inc.
Advisors: Marcel Blais, Stephan Sturm
Students: Youwei Hu, Jeremy Macaluso

The purpose of this MQP is to develop a model to predict costs and construct a bidding strategy for Cidewalk, an online advertisement platform for mobile marketing developed by Chitika, Inc.. We create algorithms to minimize the total cost of a set of Vickrey auctions for ad impressions. We develop an equation to calculate the estimated total cost of winning a set of auctions and from that we are able to obtain an optimal bidding price for the next auction.

Maintained by webmaster@wpi.edu
Last modified: Aug 04, 2016, 15:25 UTC