ENGG 390 Reports (MEM Students)

Degree Program


Year of Graduation


Faculty Advisor

Geoff G. Parker

Document Type


Publication Date

Fall 9-12-2018


Clients using the Equilend platform for post trade monitoring often experience “trade breaks” produced when small shifts in the market cause the once agreed upon terms of a trade to fall out of balance. In order to accelerate the process of fixing trade breaks, we examined historical user data and used machine learning algorithms from open source libraries to predict how those same users would fix their current and future breaks. Deliverables included background research; a clean, robust data pipeline; and a proof of concept machine learning framework. The highest performing algorithm was 98.67% accurate at predicting client break solutions.

Level of Access

Restricted: Campus/Dartmouth Community Only Access


Available to Dartmouth community via local IP address.