3 Industries Being Changed by Machine Learning
If you have spent any time around tech developers in the past five years, you’ll have noticed that one phrase is on the tip of everybody’s tongue: machine learning. What exactly is machine learning?
Simply put, it is a broad way of describing any technological system that can learn and improve its operation without being directed by a human controller.
In terms of practically developing such systems, innovators rely upon three developments:
- The proliferation and ease of harvesting of massive amounts of data
- The development of algorithms that can sort and analyze this data
- The exponential increase of computing power and storage
Practical machine learning generally involves a machine or system being able to harvest, analyze and act upon vast quantities of quantitative data. Machine learning can offer all sorts of advantages. It can allow a system to diagnose its own faults before they cause problems, simplify human/ machine interfacing, and generally make everything run a great deal more smoothly. Machine learning is making big waves in all sorts of tech-centric industries. Here are three fields where machine learning is making a mark.
How is machine learning changing the way that business water and city mains water is being delivered? It has to do with spotting faults and recognizing demand. By analyzing both sensors within itself and the data collected by users, a smart water distribution system can optimize itself without human impact.
Here is an example of how that might work. Let’s say that at midnight every night, a district in a city seems to be using an unusual amount of water. In usual circumstances, a human audit might determine the cause of this. A smart system will analyze sensors and user data to determine whether this is due to a fault or whether people are taking more baths around midnight in one district. It will then ‘learn’ from this data and decide whether more water should be allocated to the area when everybody is taking their baths.
You might have noticed that Google Translate no longer spews out clumsy garbage that makes no sense whatsoever. This is because it was quietly redesigned using machine learning principles. Machine learning allows the service to analyze learned context as well as base meaning. This allows it to pick the most appropriate translation based on what it thinks that you actually mean. Of course, this can go wrong, but it usually works well because of the sheer amount of data available to the system.
The self-driving car used to be the stuff of science fiction fantasy, but it is fast becoming a reality largely thanks to machine learning. Driving a modern car generates a huge amount of data. Your car logs where you go, how fast, how close you are to other vehicles and what kind of traffic you are likely to encounter. All of this can be used to allow self-driving cars to learn by constantly evaluating an ever-growing dataset. Theoretically, self-driving cars should grow safer the more they are used.