The student-run Data Science @ UCSB group held a project showcase on Apr. 17, to show off the projects members have been working on all year. This year, linguistics student Parker Glenn walked away with the award for Most Innovative while Dexter 2.0 won the award for Most Impactful.
The entryway of the MultiCultural Center (MCC) was packed wall-to-wall with people standing and socializing. Attire ranged from freshly ironed suits to Patagonia sweaters with crisp data science club logos. The smell of food was in the air, with a pile of sandwiches, cookies, and more on a table nearby.
“We have a group of folks who have been dedicating their evenings after class, their free time to do these projects over the past two months,” said Tim Nguyen, President of Data Science @ UCSB. “I’m just going to let their work speak for itself.”
Parker won the Most Innovative award for his practical application designed to prevent newsfeed cluttering. For example, when important developments like the Mueller report are released, multiple news outlets will cover the topic, cluttering one’s newsfeed with similar articles from each outlet. The project was also meant to filter out multiple articles of the same tone or political affiliation, allowing one to read from a diversity of perspectives to see where they differ.
The Most Impactful award was given to the Dexter 2.0, which was designed to address an important medical issue. It takes images of a blood sample at the cellular level and picks out and classifies white blood cells by one of their four types. It does so at a price much cheaper than the commercial standard which can cost a patient anywhere from $100 to $3000. The program itself has demonstrated 84 percent accuracy, which is on par with current public blood-testing software. It takes 30 minutes to build on a laptop, according to its team.
The showcase’s judges included Luca Foshcini (co-founder of data science company Evidation Health), Victor Borda (director of data science at Invoca) and Dr. Kharitonova (a computer science professor at UCSB).
During the event, they all sat in the front of the room and asked questions about each project. Often, they would ask if the students had tried a certain method and what results they had found, informing the students in the process. Other times, they asked because of a gap in their own understanding — data science is a broad field, with many frameworks and approaches to work from, so there are bits even experts don’t know.
Nearly all in attendance were from the data science club. Those who were not presenting waited to see how far their fellow club members had come and what they had managed to make in a year.
The field itself has a wide breadth of possibilities — data science is defined as any experiment or application that takes data from the past (usually a lot of it) and converts it into a model that can be used to predict future events. It is a field made more practical as computing power increases, but people of all disciplines can learn its skill sets.
Calvin told the club how proud he was to have mentored them over the last few years, and thanked them. “Today is the highlight of my UCSB career,” he said. As the crowd began to leave, the showcase’s projects were doubtless on their mind … along with the question of what next year might bring.
To read more about the rest of this year’s Data Science @ UCSB projects, click here.