Google has launched a deep learning course explaining the machine learning technique that validates so many of its services. Available through Udacity, a for-profit education site-the class is expected to take about three months to complete at a rate of six hours of work per week. It’s not for the utmost beginners if you have some background in machine learning you can tune into this course.
Through this course, participants will be introduced to TensorFlow, the in-house machine learning software that Google open sourced last November. Being an increasingly popular type of artificial intelligence, deep learning involves training artificial neural networks on lots of data. It is also the leading force behind the company’s speech recognition technology and Google Photo’s search engine.
“Deep learning is a modern take on the old idea of teaching computers, instead of programming them,” Vincent Vanhoucke, Principle Scientist at Google and technical lead on the Google Brain Team, mentioned in an Udacity blog post. “It has taken the world of machine learning by storm in recent years, and for good reason! Deep learning provides state-of-the-art results in many of the thorniest problems in computing, from machine perception and forecasting, to analytics and natural language processing.”
Mr. Vanhoucke, who is teaching the course (part of Udacity’s machine learning engineer Nanodegree program), explained that its aim is to make deep learning more accessible, and TensorFlow certainly met the needs of a wide range of users.
TensorFlow, the powerful piece of deep learning software from the Google Brain Team, will be covered by the Udacity course. And just like the TensorFlow software, the course itself is free. Udacity will also cover convolutional neural networks, recurrent neural networks, and long-short-term memory networks according to Vanhoucke.