Siemens Competition Winners: Andrew Komo and Robert Yang
by Marie Brodsky '20 for Silver Quest
It may not be immediately apparent what common thread could lie between designing a secure multi-layer system for complex auctions and developing a novel machine learning model for diagnosing lymphoma. Andrew Komo conducted his research under economics professor Dr. Lawrence Ausubel’s guidance at the University of Maryland, and proceeded to win the Grand Prize in the national Siemens Competition for math, science and technology (siemensusa). Robert Yang, a fellow Blair senior, interned at the National Institutes of Health and earned the title of Regional Finalist in the same competition (Silver Chips). The two students submitted their projects to a pool of over 1800 entries and were selected from nearly 500 semifinalists nationwide.
Over at NIH, seeking a group of students to get involved, Robert’s mentor Dr. Calvin Johnson had long wanted to develop a machine learning model to distinguish between different types of lymphoma to allow for more accurate medical diagnoses. Robert worked jointly with his team to develop the model, which was undoubtedly his favorite part. He then independently created a web server to hold information on genetic mutations of lymphoma. It was a rewarding surprise when the machine learning model performed at 100% accuracy on the 50 tested sample cases. Recognized as Regional Finalist the previous year for a team project on neuroscience (siemens-foundation), Robert says that this year’s project allowed him to take much more of a leadership role than he had the year before.
For Andrew, the project began with his mentor proposing the development of a system for secure proxy bidding in an auction. Building off this idea, Andrew decided to challenge himself and design a more complex protocol for multi-item auctions. His project was almost completely self-guided from there, and even his research paper was written independently. Andrew used occasional meetings with his mentor to ensure that his current ideas were viable to be applied in a real auction. His developed auction system was composed of even more layers of complexity than he had initially realized; “I had to view the auction as a whole, and then modify and synthesize my preliminary ideas so that they would work together and have no security vulnerabilities.” Overall, Andrew found it “very satisfying to use [his] computer science skills and knowledge to fill a void.”
Andrew Komo wins $100,000 Grand Prize
Senior Andrew Komo was the individual winner of the 2017 Siemens Competition, winning a $100,000 scholarship (siemensusa).
The staggering 410 hours that Andrew estimates he put into his project included months of self-studying on cryptography and auction theory at home and seven 40-hour workweeks during the summer at Power Auctions LLC. Robert, similarly, researched immunology and computer science before beginning his internship, and consequently spent almost the entire summer as well as the start of the school year working on his project.
Despite difficulties along the research process, both awardees found great interest in their work, which have served as an opportunity to immerse themselves completely in experimentation and realize the particular aspects that drive their passion for their respective fields of interest. Robert says he now feels confident in continuing to pursue the intersection of computer science and biology rather than choosing one of the two, and Andrew feels that he has crystallized his interest in the intersection of computer science and economics. The competition itself, as Andrew said, was “very validating . . . and the finals were a lot of fun.”
For students who have yet to begin projects in research, the two have some advice: Finding a mentor may not be the easiest task; Robert successfully reached out to his mentor after taking interest in his publications. As for the competition aspect, Andrew recommends using the summer to write the research paper and to complete the questions for the competition.