I’m a rising junior at Simmons College in Boston, Massachusetts, pursuing a B.S. in Computer Science. This summer, I’m working under Dr. Guo-Jun Qi on the research of generative adversarial nets within the field of machine learning. Specifically, we are looking at Loss-Sensitive GANs, which work to more effectively and efficiently discern, through the discriminative model, between “real” data inputs and “fake” data generated by the generative model during the training process. The loss-sensitivity factor adds precision to the GAN in that it helps prevent more accurate generated data from being categorized as negative by the discriminator, and impacting the learning process.
REU Symposium Presentation Powerpoint