Why We Created Conf.Startup.ML


An estimated 11 million people are involved in the software industry today (according to IDC). They write complicated, brittle software using a declarative approach that often fails to recognize the complexity of the real world and the intricacy of user behavior.

The future of software engineering is not declarative, it’s probabilistic!  Conf.Startup.ML is a one-of-a-kind event that brings together the leading machine learning engineers from industry and academia to kick off the transition away from the declarative era of software development.  Organized by Startup.ML, Next.ML features leading machine learning researchers from Facebook, Université de Montréal LISA Lab, Stanford, and industry innovators including indico, Domino Data Lab, MapR, GraphLab, and H2O.ai.

Facebook’s Soumith Chintala will be leading a workshop on video classification using convolutional neural networks, the same algorithm that powered this recent Stanford research project on image annotation.  Soumith will take attendees through the processes of training deep neural nets using GPUs on AWS. Other workshops will cover image analysis, sentiment analysis, probabilistic programming and the new scientific compute language-Julia.

Conf is not a series of research papers turned into powerpoint presentations, rather it is a chance to equip yourself with the intuition and necessary skills to solve problems through machine learning.