Project-based learning opportunities come in all forms at MIT, as Melanie Chen discovered during her internship at Lincoln Laboratory this year. A computer science major, she served as a teaching assistant, curriculum developer, and mentor to high school students participating in Cog*Works, part of the Beaver Works Summer Institute. Now, finishing up her fall sophomore semester, Chen is finding plenty of opportunities to apply the lessons she has learned from her hands-on experience teaching others.
“One of the greatest skills I’ve learned is effective communication,” she notes. “Whether it's students, peers, colleagues, or mentors, I've learned what it takes to be able to create a trusting relationship, so that we can effectively work on a project of this scale together.”
Cog*Works, a summer course offered to talented rising-senior high school students, gives interns hands-on experience developing technologies used in artificial intelligence applications. Created by Ryan Soklaski, a technical staff member at Lincoln Laboratory, it promotes project-based learning using Python and other open-source technologies, introducing students to powerful and accessible tools for customizing their own cognitive assistants.
In the months leading up to her internship, Chen helped develop the curriculum for Cog*Works, and was responsible for creating course materials that would teach students how to “hack” the Amazon Echo Dot, allowing them to customize the on-board cognitive assistant, Alexa, using their own code and hardware. At the beginning of July, she took on the role of teaching assistant throughout the intensive, full-time, four-week course. She and the other course instructors worked with 28 high school students from around the country, introducing them to audio processing, computer vision, and natural language processing techniques. Students created their own song-recognition algorithms, trained neural networks, and coded search engines from scratch. In one project, students equipped Alexa with a camera and access to a neural network capable of performing state-of-the-art face detection and recognition. Alexa could thus “recognize” who is speaking to it — a capability that is expected of next-generation cognitive assistants.
“Cog*Works empowered its students,” Chen stated. “It taught them that they are capable of creating impressive tech, and that this creative process is exhilarating.”
As impactful as the experience has been for the Cog*Works interns, it also left a lasting impression on Chen. “I have some prior teaching experience, but I've never really taught students who are this close to my age range,” she said. “The challenge was to project myself as what I would want to see in a mentor.”
The experience also jump-started Chen’s facility with machine learning techniques, which are covered in 6.867 (Machine Learning), a course she is currently taking this semester. “Being able to prepare myself such that the students could actually ask me questions, and I’d be able to understand the material to an extent that I could answer all the questions they have, was challenging in the weeks leading up to the course.” But Chen persevered, and is now applying her newfound skills to her coursework and research in the Spoken Language Systems Group in the Computer Science and Artificial Intelligence Laboratory. She plans to pursue a master’s degree with a focus in machine learning. She is also the MIT Animation Group public relations director, a radio show host at WMBR, and an MIT Arts Scholar.
Submitted by: Lincoln Laboratory | Video by: Meg Rosenburg/MIT Video Productions | 3 min, 54 sec
