Source: Major League Hacking Twitter
We just completed a pilot workshop with Major League Hacking and it was a huge success! Brent recaps the experience.
As part of our mission to make machine learning education more accessible, we have partnered with Major League Hacking (MLH) to deliver an interactive workshop through their Localhost program.
MLH Localhost rocks
Localhost is essentially an event-in-the-box that is shared with MLH-trained instructors that can run workshops at schools across the nation. It's a great way to get student hackers started on their path to explore machine learning in a fun and interactive environment.
We've had a great time working with the MLH team to develop interesting (and entertaining) use cases to present at the Nexosis workshop. In this case machine learning was used to predict what characters said various quotes from the Back to the Future trilogy. Our product really seemed to resonate with the MLH team. They came up with a very surprising amount of content with minimal input from our team, which served to demonstrate how easy we really do make it to get up and running with a machine learning project.
Our product really seemed to resonate with the MLH team. They came up with a very surprising amount of content with minimal input from our team, which served to demonstrate how easy we really do make it to get up and running with a machine learning project.
We recently had a chance to pilot the workshop at MLH's New York headquarters. There were about twenty people in attendance with our own Guy Royse on site to provide support. It was great to see folks engaging with our product and also how smoothly the Localhost content was delivered. It was a very positive event, and we're excited to see it repeated for more students.
We ran into some speed bumps along the way and will need to spend some time retooling the workshop to scale more effectively for more attendees. One of the challenges with an event like this is that our automated machine learning engine favors accuracy over speed, so it is difficult to guarantee that a large batch of model-building sessions will finish within the time allotted for the workshop. What we do guarantee, however, is that each model we generate will be the best that we can produce.
It was very encouraging to see students from that pilot workshop using our API and dashboard over the following days to repeat the MLH exercise and to complete some of the tutorials we've posted here on our website. We hope that all who attended had a good time and learned something new, and we're eager to share this experience with more hackers in the months to come!