What is machine learning and what can it do?
You step into the tennis section of your local sports store looking for a new racket when a notification alerts you of a coupon on Wilson products. Something must have picked up your location along with your online shopping history to know to throw you a bone at that exact instant. That something is machine learning.
Machine learning is a technique by which a computer learns to compute a task without explicit instruction. It is used to train computers by using algorithms that iteratively learn from data and enable them to find insight. The core subarea of artificial intelligence, it can be used to teach computers to complete a task, make accurate predictions or behave intelligently. The learning that is being done is always based on observations or data, such as examples, direct experience or instruction, according to Rob Schapire, former computer science professor at Princeton. Schapire said, in general, machine learning is about learning to do better in the future based on what was experienced in the past.
“Often we have a specific task in mind, such as spam filtering,” Schapire said. “But rather than program the computer to solve the task directly, in machine learning, we seek methods by which the computer will come up with its own program based on examples that we provide.”
Schapire, inventor of the AdaBoost algorithm, gives several examples of outcomes enabled by machine learning, including optical character recognition, face detection, spam filtering, topic spotting, spoken language understanding, medical diagnosis, customer segmentation and fraud detection.
Enabling a smarter IoT
In the past decade, machine learning has helped support self-driving cars, speech recognition, web search optimization and an improved understanding of the human genome. According to Stanford, machine learning is so pervasive today that you probably use it dozens of times per day without knowing it. Machine learning is not a new concept, but it is being used in new ways to enable the “internet of things.”
Data is an integral part of IoT, but the incredibly large amounts being gathered can be difficult to search through. Decisions of what data to keep, ignore and what to forward to a centralized authority will need to be made instantly with local information and knowledge. Most IoT endpoints are expected to be limited in capabilities due to size, cost and power requirements, and will need companion computing that is either embedded in the larger system or in an IoT gateway, according to Moor Insights and Strategy.
All of that digestible data is being used throughout a host of industries for IoT solutions. SAS Insights provides a number of industries using machine learning and IoT to optimize processes:
Health care
Wearable devices and sensors can use data to assess a patient’s health in real time. The technology can also help medical experts analyze data to identify trends or red flags that may lead to improved diagnoses and treatment.
Marketing and sales
Websites recommending items you might like based on previous purchases are using machine learning to analyze buying history and promote other items you’d be interested in.
Oil and gas
Finding new energy sources, analyzing minerals in the ground, predicting refinery sensor failure, etc.
Transportation
Analyzing data to identify patterns and trends is key to the transportation industry, which relies on making routes more efficient and predicting potential problems to increase profitability.
By 2020, MI&S believes machine learning will exist in a large number of solutions and will account for a great deal of the innovation in the IoT world with companies like IBM, with Watson IoT; or Google and TensorFlow leading the way.
