What we're about
The DC Deep Learning Working Group (DC DLWG) is for those interested in becoming proficient with Deep Learning concepts and coding. Our approach is highly collaborative and we learn through discussion rather than by just listening to lectures. We use journal papers, video lectures and Deep Learning courses to drive our discussions and provide coding exercises. Occasionally group members present special related topics helpful to the group’s understanding. Our group grew out of the Machine Learning Journal Club, led by Don Vetal, which focused on studying academic papers related to machine learning, optimization and data science and pursued knowledge of the real guts of the algorithms. We continue keeping to that spirit, but now focus on coding as well.
We generally meet weekly. The meeting format typically alternates between lecture/paper discussions and lab sessions where we review code. In our lecture sessions we discuss and gain a better understanding of course lectures. In our lab sessions, we walk methodically through code from course assignments. We intend to expand our projects beyond the course material, based on the interests of the group.
We welcome all new members and participants, regardless of experience level, who are excited about rolling up their sleeves to dig into Deep Learning. We use Python and Google’s TensorFlow framework. We assume participants are comfortable enough with programming to pick up Python on their own. In August 2016 we finished covering the Udacity course on Deep Learning and TensorFlow (https://www.udacity.com/course/deep-learning--ud730). It would be highly beneficial, though not required, to review that material, as we proceed in September 2016 to follow Stanford’s CS224d Deep Learning for Natural Language Processing (http://cs224d.stanford.edu/).