Three Types of Learning That Artificial Intelligence (AI) Does

If you are just starting your journey into the most hottest field right now -Machine Learning, then you must have heard of these confusing words — Deep learning, Reinforcement learning, Supervised learning, Unsupervised learning.

Oh man! Here come the old topics that I had been learning ever since I started machine learning. But the thing that I found with new comers is that they often get confused by definitions of these terms. They have questions like below:

  • Is machine learning and artificial intelligence the same?

  • What is the difference between machine learning and deep learning?

  • How is supervised, unsupervised learning related to machine learning and deep learning?

After you complete reading this post to the end, I am pretty sure that you will never get pop up questions like the ones above.

Artificial Intelligence and Machine Learning

First of all, let me tell you this — AI and ML are not the same. We use machine learning to give intelligence to the machines(computers). There are two ways by which we can give artificial intelligence to machines: By Rule-Based Programming or by using Machine Learning.

In machine learning, we let the machine learn by itself. Ok, let me make this clear: How do you learn for your exams? You read the textbooks provided by your school/college and learn from the contents in the textbooks and thus you can perform well in your exams. Similarly, you provide the machines(computers) with a dataset and the machine learns from the dataset and does the prediction. .

Okay, just think like this: you are a machine learning research scientist at a healthcare startup. You are asked to create a machine learning model to predict whether a person has leukemia or not by analyzing his/her blood sample.

So, generally what you as a research scientist will do is that you will train(teach) your model by giving it a huge dataset which contains images of blood cells(previously analyzed by human doctors) with corresponding label as leukemic or non-leukemic.

After training, the model must have understood the patterns for leukemic and non-leukemic cells. Now if you give an image of blood cell collected from a new person, it should be able to predict the right answer(leukemic or non-leukemic).

Now the machine is intelligent enough(like an expert doctor) to identify a leukemic cell from the non-leukemic.

Deep Learning Deep learning is similar to or we can call it as a subset of machine learning. The method for deep learning is similar to machine learning(we let the machine learn by itself) but there are a few differences. Some of them are:

  1. Algorithms used in deep learning are generally inspired from human neural networks.

  2. Deep learning requires huge datasets and computational power(you guessed it right -GPU’s) than machine learning.

So, when ever you use an algorithm which has a word -Neural Network, or when ever your algorithm requires a huge dataset to learn -Yo man! You are using the hottest topic of the present world, deep learning.

Reinforcement Learning

The general overview about this topic that I can give you is that -Rewards and Punishments.

Oh! What the heck is that? Ok, let me ask you a question.

How did you learn to walk in your childhood? Through Trial and Error, right?

You will try your maximum to walk without falling -falling down while walking is a punishment and walking without falling is a reward.

So, what will you do? You will try to minimize the punishment(falling down) and maximize the reward(walk without falling down).

Have you ever heard of Google’s Deepmind?

Deepmind created a robot named AlphaGo which uses this technique(reinforcement learning) to play the most difficult board game — Go.

AlphaGo beat the world champion in Go. Sounds amazing, right? Get ready to get really amazed.

Deepmind created another version of AlphaGo named AlphaGo Zero which beat AlphaGo hands down(100–0 or something like that). The field is just in it’s budding stage, this is the right time to play with this cool stuff. Just keep exploring this wonderful field of machine learning.

Artificial intelligence/Machine learning What is machine learning?

Machine-learning algorithms use statistics to find patterns in massive* amounts of data. And data, here, encompasses a lot of things—numbers, words, images, clicks, what have you. If it can be digitally stored, it can be fed into a machine-learning algorithm.

Machine learning is the process that powers many of the services we use today—recommendation systems like those on Netflix, YouTube, and Spotify; search engines like Google and Baidu; social-media feeds like Facebook and Twitter; voice assistants like Siri and Alexa. The list goes on.

In all of these instances, each platform is collecting as much data about you as possible—what genres you like watching, what links you are clicking, which statuses you are reacting to—and using machine learning to make a highly educated guess about what you might want next. Or, in the case of a voice assistant, about which words match best with the funny sounds coming out of your mouth.

Frankly, this process is quite basic: find the pattern, apply the pattern. But it pretty much runs the world. That’s in big part thanks to an invention in 1986, courtesy of Geoffrey Hinton, today known as the father of deep learning. What is deep learning? Deep learning is machine learning on steroids: it uses a technique that gives machines an enhanced ability to find—and amplify—even the smallest patterns. This technique is called a deep neural network—deep because it has many, many layers of simple computational nodes that work together to munch through data and deliver a final result in the form of the prediction. What are neural networks? Neural networks were vaguely inspired by the inner workings of the human brain. The nodes are sort of like neurons, and the network is sort of like the brain itself. (For the researchers among you who are cringing at this comparison: Stop pooh-poohing the analogy. It’s a good analogy.) But Hinton published his breakthrough paper at a time when neural nets had fallen out of fashion. No one really knew how to train them, so they weren’t producing good results. It took nearly 30 years for the technique to make a comeback. And boy, did it make a comeback. What is supervised learning?

One last thing you need to know: machine (and deep) learning comes in three flavors: supervised, unsupervised, and reinforcement. In supervised learning, the most prevalent, the data is labeled to tell the machine exactly what patterns it should look for. Think of it as something like a sniffer dog that will hunt down targets once it knows the scent it’s after. That’s what you’re doing when you press play on a Netflix show—you’re telling the algorithm to find similar shows.

What is unsupervised learning?

In unsupervised learning, the data has no labels. The machine just looks for whatever patterns it can find. This is like letting a dog smell tons of different objects and sorting them into groups with similar smells. Unsupervised techniques aren’t as popular because they have less obvious applications. Interestingly, they have gained traction in cybersecurity. What is reinforcement learning?

Lastly, we have reinforcement learning, the latest frontier of machine learning. A reinforcement algorithm learns by trial and error to achieve a clear objective. It tries out lots of different things and is rewarded or penalized depending on whether its behaviors help or hinder it from reaching its objective. This is like giving and withholding treats when teaching a dog a new trick. Reinforcement learning is the basis of Google’s AlphaGo, the program that famously beat the best human players in the complex game of Go.

That’s it. That's machine learning. Now check out the flowchart above for a final recap.

2 views0 comments