Ever wondered how Netflix is able to provide you highly relevant video suggestions? Or how Amazon is able to suggest new products to you, even though you never browsed through them in its catalogues? Or better still, social media sites such as Facebook and LinkedIn are able to offer “people that you may know” suggestions so accurately? Yes, these are all examples of machine learning that you might come across during your day to day life. So, what exactly is machine learning, and why has it suddenly become so important?
As defined by Wikipedia, “Machine learning (ML) is a field of computer science that gives computers the ability to learn without being explicitly programmed.”
When I was growing up in the 80s/90s, machine learning was just a buzzword. Coupled with artificial intelligence (AI), it was often the topic of discussion, but always in the context of sci-fi and definitely something beyond our abilities then. The primary reason for this, was the fact that computer science was still grappling with the limitations of scale – of data, hardware and computing power. With the advent of the 21st century, and rapid strides in greatly reducing the constraints mentioned above, our computer systems became very powerful. Along with this, our ability to mine data grew exponentially. All this resulted in enormous computing power being handed down to the masses.
This is best explained by the simple, yet lucid Howard graph –
Machine learning is heavily dependent on quality and volume of data that powers it. With the advent of big data, both the amount of data available and the ability of our machines to process it has increased tremendously. All of these mean, it is possible now to quickly and automatically produce models that can analyse bigger, more complex data and deliver faster, more accurate results – on very large scales altogether. And by building precise models, an organization has better chances of identifying profitable opportunities, or avoiding unknown risks. The best part is – most of this happens in real time – thus, helping companies gain a distinct competitive advantage.
Most industries have realized the power and value offered by machine learning techniques. Here’s a list of major industries and how they are using machine learning.
Government
Agencies and organizations are mining voluminous, multiple and disparate data sources from government systems for insights to increase efficiency and detect and minimize fraud. It has been recently reported that Aadhaar based services, which rely heavily on ML systems, helped save the government over $9 billion by eliminating fraud from beneficiary listings.
Healthcare
With the advent of wearable devices and sensors, medical experts are analysing data to identify trends or red flags in patients, resulting in improved diagnosis and treatment. Microsoft’s Project Hanover is collaborating with the Knight Cancer Institute to develop AI technology for cancer precision treatment, with a current focus on personalizing drug combinations.
Transportation
Perhaps the most widely discussed application of ML is in the field of transportation with Google talking incessantly about self-driven cars and Elon Musk’s SpaceX successfully launching unmanned rockets into space. With far reaching applications, such as tracking congestion, preventing crashes and collisions and optimizing commute times, it is widely believed that ML will completely disrupt transportation as we know today.
Marketing & Sales
Major online stores such as Flipkart, Amazon, Myntra are already using ML to offer buying recommendations based upon your previous purchases or catalogue browsing behaviour. This ability to capture consumer shopping data, analyse it, and personalize the shopping experience will form the future of retail.
Financial Services
Banking and financial institutes are using ML for two key purposes – customer insights and preventing fraud. Insights can help in identifying investment opportunities, especially in case of high-net-worth individuals. Cybersurveillance, facilitated by big data sources, can pinpoint warning signs of fraud. JPMorgan, Bank of America and Morgan Stanley are already developing automated investment advisors or “fin-bots”, powered by ML, for automating investment predictions for customers.
All in all, machine learning and its applications will have an ever-increasing impact on our lives, in spheres and areas that we can hardly imagine as of now.
Question is, are we as businesses prepared for it? We cannot be prepared – only at our own peril.
