Machine learning is something that everyone seems to be talking about this days. Especially after everyday folks see the headline ” Facebook Shuts Down Robots After They Invent Their Language.”. Well before discussing anything further, let’s talk what exactly is machine learning is. Quoting Tim Mitchell,
“A computer program is said to learn from experience E concerning some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E.”
In this definition:
- Task T is what machine is seeking to improve. It can be something like prediction, classification, clustering, etc.
- Experience E can be training data or input data through which the machine tries to learn.
- Performance P can be some factor, like improvements in accuracy or new skills that the machine was previously unaware of. While to everyday folks it would seem like that machine learning is a new idea the truth is that it started as far back as the 1950s when computer scientists figured out how to teach a computer to play checkers.
Machine learning and artificial intelligence have applications in almost every field possible. A few includes
1) Machines that can See:
Courtesy of machine learning algorithms, it’s relatively easy to write an algorithm that can recognize characteristics in a group of pictures and categorize them appropriately. For, e.g. an algorithm has been written that can detect cancer more accurately than the best pathologists, freeing up the doctors time to make the treatment decisions more quickly and accurately. Also, the fact that computers can visualize the world around them is the key to driverless cars. And this innovation alone could revolutionize many different business models, from the supply chain and delivery to personal transportation.
2) Machines that can Write:
While it would take a hell lot of computers together to produce something of the Shakespearean quality, but computers are getting a lot better at creative writing. To prove it a trained computer was asked to write captions for a bunch of photos. In its first iteration, the human readers thought that the computer-generated captions were better than the human-generated 1 out of 4 times. This has vast implications in the fields of Data entry and classification tasks that previously required extensive human intervention. If a computer can recognize something be it an image, file, photo or even a doc file and describe it accurately, there could be many uses for such automation.
While these applications lie on the more technological side of things, there are some applications which seem more mainstream or consumer friendly.
3) Machines that can speak :
Cortana, Siri, Google Assistant are some of the favorite examples of digital assistants. Your work is simple, activate them by saying a catchphrase and ask your question like “what’s the weather going to be today” or “Send an email to xyz person”.For answering, your assistant looks out for the information, recalls your related queries, or send a command to other resources (like phone app or the weather app ) to collect the required info. Machine learning is an essential part of these assistants as they respond to your queries on the basis of your previous interaction with them. Later, this set of data is utilized to render the results that are tailored to your specific preferences.
4) Email Spam and Malware:
Almost every one of us has at some point in our lives have received spam mail and it annoys us to death. Lucky for us most of the email clients today include spam and malware filtering. To make sure that these filters are continuously updated and are up to date they are powered by various machine learning algorithms such as Decision Tree Induction, Multi-Layer Perceptron. If you go by the rule-based spam filtering, there is a very high chance that it would fail to track the latest trickery adopted by the trickster.
The thing is that each piece of code that is written is around 90-98% similar to the previous version. The machine learning powered system can understand the coding pattern. Therefore, they can detect new malware with 2-10% variation much easily.
Also Read: 15 Awesome WhatsApp Tricks You Never Knew
5) Online Customer Query and Support:
Nowadays whenever we visit a website, there is a little question mark or message icon in the bottom left of the site. That is the customer support plugin. While for a small startup they might have a real person in the backend to answer all the queries, but for large business houses, it is just not possible due to the sheer number of questions they receive in a single day.
So it is a very high chance that you are talking to a chatbot. These chatbots extract the required information from the website and present it to the customer. While you get your query sorted these chatbots learn and became smarter with time. They tend to understand the user queries a little better and serve them with better and possibly correct answers, which is only possible due to its underlying machine learning algorithms.
6) Online Product Recommendations:
Imagine a scenario that you bought a specific item from an e-commerce website. What you will notice that after a few days you will start receiving emails regarding similar items, discount codes and coupons for further shopping. Also, the next time you visit the website, you will notice items are displayed based on your shopping history. Indeed, this refines the shopping experience for you and machine learning algorithms are the one to thank here. On the basis of your past behavior on the website like a wishlist, prior purchases, items liked or added to the cart, the site learns about you and makes the appropriate product recommendations.
These are but some applications of machine learning in our lives. There are countless more that if discussed would make this article unbearably long.
The thing about machine learning or artificial intelligence for that matter is that we don’t know everything about it just yet. We learn something new about it as more and more research comes to light. So are we close to real artificial intelligence? Some believe that’s the wrong question. They feel that a computer will never “think” in the way that a human brain does, and that comparing the computational analysis of a network to the mechanics of the human mind is like comparing apples and oranges.
Regardless, computers’ abilities to understand, and interact with the world around them is growing at an unprecedented rate. And as the quantities of data continue to grow exponentially, so will our computers’ ability to process and analyze — and learn from — that data grow and expand.
I am a sophomore at IIT Roorkee currently pursuing my B.tech in Metallurgical and Materials Engineering.