Applications of Machine Learning in Real World
Applications of Machine Learning:
1. Software: Machine learning is widely used in software to enable an improved experience for the user. With some packages, the software is learning about the user’s behaviour after its first use. After the software has been in use for some time, it predicts what the user wants to do.
2. Spam Detection: For all the junk mail that gets caught, there’s a good chance a Bayesian classification filter is doing the work to catch it. Since the early days of SpamAssassin to Google’s work in Google Mail, there’s been some form of learning to figure out whether a message is good or bad.
Spam detection is one of the classic uses of machine learning, and over time the algorithms have gotten better and better. Think about the e-mail program that you use. When it sees a message it thinks is junk, it asks you to confirm whether it is junk or isn’t. If you decide that the message is spam, the system learns from that message and the experience. Future messages will, hopefully, be treated correctly from then on.
3. Voice Recognition: Apple’s Siri service which is on many iOS devices is another example of software machine learning. You ask Siri a question, and it works out what you want to do. The result might be sending a tweet or a text message or setting a calendar appointment. If Siri can’t work out what you’re asking of it, it performs a Google search on the phrase you said. Siri is an impressive service that uses a device and cloud-based statistical model to analyze your phrase and the order of the words in it to come up with a resulting action for the device to perform.
4. Stock Trading: There are lots of platforms that aim to help users make better stock trades.
These platforms have to do a large amount of analysis and computation to make recommendations. From a machine learning perspective, decisions are being made for you on whether to buy or sell a stock at the current price. It takes into account the historical opening and closing prices and the buy and sells volumes of that stock.
5. Robotics: Using machine learning, robots can acquire skills or learn to adapt to the environment in which they are working. Robots can acquire skills such as object placement, grasping objects, and locomotion skills through either automated learning or learning via human intervention. With the increasing amount of sensors within robotics, other algorithms could be employed outside of the robot for further analysis.
6. Medicine and Healthcare: The race is on for machine learning to be used in healthcare analytics. Several startups are looking at the advantages of using machine learning with Big Data to provide healthcare professionals with better-informed data to enable them to make better decisions.
IBM’s famed Watson supercomputer once used to win the television quiz program Jeopardy against two human contestants is being used to help doctors. Using Watson as a service on the cloud, doctors can access learning on millions of pages of medical research and hundreds of thousands of pieces of information on medical evidence.
7. Advertising: For as long as products have been manufactured and services have been offered,
companies have been trying to influence people to buy their products. Since 1995, the Internet has given marketers the chance to advertise directly to our screens without needing television or large print campaigns. Remember the thought of cookies being on our computers with the potential to track us? The race to disable cookies from browsers and control who saw our habits was big news at the time.
8. Retail and E-Commerce:
Machine learning is heavily used in retail, both in e-commerce and brick-and-mortar retail. At a high level, the obvious use case is the loyalty card. Retailers that issue loyalty cards often struggle to make sense of the data that’s coming back to them.
9. Gaming Analytics: We’ve already established that checkers is a good candidate for machine learning. Do you remember those old chess computer games with real plastic pieces? The human player made a move and then the computer made a move. Well, that’s a case of machine learning planning algorithms in action. Fast-forward a few decades to today when the console market is pumping out analytics data every time you play your favourite game.
Microsoft has spent time studying the data from Halo 3 to see how players perform on certain levels and also to figure out when players are using cheats. Fixes have been created based on the analysis of data coming back from the consoles.
10. Internet of Things: Connected devices that can collate all manner of data are sprouting up all over the place. Device-to-device communication is hardly new, but it hadn’t hit the public mind until fairly recently. With the low cost of manufacture and distribution, now devices are used in the home just as much as they are in the industry.
Uses include home automation, shopping, and smart meters for measuring energy consumption. These things are in their infancy, and there’s still a lot of concern about the security aspects of these devices. In the same way, mobile device location is a concern, companies can pinpoint devices by their unique IDs and eventually, associate them with a user.