AI has been part of our imaginations long before it hit the news. Visions of research labs and computer scientists rallied around the latest self-aware technology has filled not only our nightmares, but the afternoon cinemas. Birthed in 1956, AI has alternately been heralded as the key to our civilization’s brightest future and the preposterous fantasies of geeks. Until 2012, it was a bit of both.
Today’s AI is applied in varying ways and degrees. Machine learning, a field of AI, is the practice of using algorithms to parse data, learn from it, and then make a determination or prediction about something in the world. So rather than hand-coding software routines with a specific set of instructions to accomplish a particular task, the machine is trained using large amounts of data and algorithms that give it the ability to learn how to perform the task.
Machine learning came directly from minds of the early AI crowd, and the algorithmic approaches over the years included decision tree learning, inductive logic programming. clustering, reinforcement learning, and Bayesian networks, among others.
Deep learning is a technique for implementing machine learning. It was inspired by the structure and function of the brain, namely the interconnecting of many neurons. Artificial Neural Networks (ANNs) are algorithms that mimic the biological structure of the brain. In ANNs, there are neurons that have discrete layers and connections to other neurons. Each layer picks out a specific feature to learn, such as curves and edges in image recognition. It’s this layering that gives deep learning its name. Depth is created by using multiple layers as opposed to a single layer.
Deep learning has enabled many practical applications of machine learning and by extension the overall field of AI. Deep Learning breaks down tasks in ways that makes all kinds of machine assists seem beyond possible. No idea is too big. Driverless cars, better preventive healthcare, even better music recommendations are either now or soon to come.
The current deep learning hype tends to be that we have machinery, that given enough data and enough training time, is able to learn on its own. This of course is either an exaggeration of what the state-of-the-art is capable of or an over simplification of the actual practice of deep learning. It has over the past few years given rise to a massive collection of ideas and techniques that were previously either unknown or known to be indefensible.
Deep learning today goes beyond just multi-level artificial neurons, but instead is a collection of techniques and methods that are used to building unique architectures.
We are looking to the future where we hope to have a more solid foundation to be able to build our learning machines with greater predictability of their capabilities.
Melody K. Smith
Sponsored by Access Innovations, the world leader in taxonomies, metadata, and semantic enrichment to make your content findable.