Practical applications of deep learning

Deep learning has received a great deal of attention lately due to its successes in tackling machine learning tasks that have been notoriously difficult to solve with conventional methods. Using sophisticated neural networks to teach computers to think more like a human has allowed machines to catch up with or even surpass human performance on many tasks that humans once held a seemingly insurmountable lead. Perhaps more importantly, even if humans still perform better at certain tasks, the upsides of machine learning—workers that never tire, never get bored, and require no salary—turn even imperfect automatons into useful tools for many tasks.

Unfortunately, for those of us working outside of large technology companies and research organizations, it is not always easy to take advantage of deep learning methods. Training a deep learning model generally requires not only state-of-the-art computing hardware but also large volumes...