About
Jared outside of Peter Shield's library at the University of California, Davis. Image edited using neural style transfer.
Jared Yu is a graduate student studying for an online M.S. in Data Science at Johns Hopkins University. His background is a B.S. in Statistics from the University of California, Davis. The expected date of graduation is around May 2021.
Jared is driven to discover the insights into problems that are important for them to be solved. By working hard to break problems down into their component pieces, it becomes possible for him to gain enough clarity to the point that a vision of the solution can begin to appear. This difficult struggle to know why a problem exists and how it can be fixed is what brings joy and satisfaction to his work.
His determination to solve such problems can be seen in the example of a feedforward neural network tutorial that he had developed in graduate school. Through the reinterpretation of the abstract textbook example as a linear regression problem utilizing matrices, the big picture became much more obvious to others (including himself) who were previously having a difficult time with the concept (6000+ views). To those unaware, feedforward neural networks are the basics of deep learning (artificial intelligence) and at first the logic is counterintuitive to what most people may have previously learned.
Another example of his passion for understanding the basic truths within a problem is in a presentation on Kernel Density Estimation (K.D.E.). The concept of K.D.E. is simple enough that it can be understood as a continuous form of a histogram, but a kernel function is included to help estimate the true density of a variable at any given point. The interesting portion of this method comes from deriving the closed-form solution for a seemingly simple task. With his experience in statistics as an undergraduate, he was able to make the objective of calculating complex cross terms for two Gaussian densities into a single straightforward step. The feedback from his peers in this presentation were positive, calling it a “gold standard” and more helpful than anything else found on YouTube.
Jared is seeking internship opportunities as he continues on the path of completing his M.S. in Data Science. Afterwards, he will be interested in any role related to data analytics (e.g., data science, business analysis, etc.). Reach out to qzyu999@gmail.com to connect!