How Machines Are Able To Learn Just Like You

Artificial Intelligence.

Two words which may invoke ideas of a Terminator-type future where machines destroy and take over the world. And while it may be scary that computers can simulate how the human brain works and learn from their mistakes, I assure you that the present impacts of AI have been much more positive than the stories depicted in movies.

In fact, rather than being the ultimate death of us all”, AI has been helping and growing so many industries across the world. It makes it possible for machines to learn from experience, adjust to new inputs and perform human-like tasks thanks to deep learning and natural language processing. Using these technologies, computers can be trained to accomplish specific tasks by processing large amounts of data and recognizing patterns in the data, just like humans!

And the best part is, the possibilities of applications of AI are endless, ranging anywhere from building an algorithm to become the best Fortnite player in the world to being able to detect if a patient has skin cancer. And while I may not approve of the former, it would be SICK if AI can help patients of some of the biggest diseases in the world.

But how does all of this work? How can AI power self-driving cars or beat world renowned chess players?

Machine Learning — the brains behind AI

There are actually many different approaches to getting machines to learn, from using basic decision trees to clustering to layers of artificial neural networks, depending on what task you’re trying to accomplish and the type and amount of data that you have available. That being said, all of these fall into the three main types of machine learning:

  1. Supervised Learning
  2. Unsupervised Learning
  3. Reinforcement Learning

Supervised Learning

Supervised learning is like teaching a child with the help of flashcards

Just like how flashcards have data in the form of examples with labels, we can feed a learning algorithm example-label pairs one by one. Through this method, the algorithm will predict the label for each example, and receive feedback as to whether it predicted the right answer or not — much like what I do when I’m using flashcards to study for my next French test.

Essentially, over time, the supervised learning algorithm will be able to observe a new, never-before-seen example and predict a good label for it.

Unsupervised Learning

Similarly, in unsupervised learning, the algorithm would be fed a lot of data and be given the tools to understand the properties of the data. From there, it can learn to group, cluster, and/or organize the data in a way such that a human (or other algorithm) can come in and make sense of what the newly organized data means.

What makes unsupervised learning such an interesting area is that an overwhelming majority of data in this world is unlabeled. And while it would be a pain to have to sort through all of it ourselves, having intelligent algorithms that can take our terabytes and terabytes of unlabeled data and make sense of it is a huge source of potential profit for many industries!!

Reinforcement Learning

Place a reinforcement learning algorithm into any environment and it will make a lot of mistakes in the beginning. Just like how one would teach a dog a new trick, a reinforcement algorithm would receive a “treat” or positive signal for every good behaviour while getting “no treat” or a negative signal when it performs bad behaviours. Thus over time, our learning algorithm learns to make less mistakes than it used to.

For any reinforcement learning problem, we need an agent and an environment as well as a way to connect the two through a feedback loop. To connect the agent to the environment, we give it a set of actions that it can take that affect the environment. To connect the environment to the agent, we have it continually issue two signals to the agent: an updated state and a “treat”/reward (the reinforcement signal for its behavior).

Let’s look at an example to understand this:

Using reinforcement learning we can train algorithms to play games like Super Mario

In the game of Mario, our agent is our learning algorithm and our environment is the game. Our agent has a set of actions, which will be the button states. Our updated state will be each game frame as time passes and our “treat” signal will be the change in score. And with all these components together, we would be able to set up an algorithm that will eventually learn how to play Mario!

Applications of AI:

Marketing

Netflix uses AI to recommend shows that you might enjoy watching

Companies are now able to provide highly accurate predictive technology based on customer’s search history, reactions, etc. It examines millions of records to suggest products that you might like based on your previous actions and choices. And, as the data set grows, this technology is getting smarter and smarter every day.

Medicine

Whether its detecting lung cancer, assessing the health of your heart, or predicting the impact of a certain drug on the human body we might be able to live longer as a result of AI. The more we digitize and unify our medical data, the more we can use this tech to help scientists find valuable patterns that can be used to make accurate, cost-effective decisions to improve our healthcare industry. And if that’s not impressive enough, let’s take a look at another industry AI has been affecting:

Agriculture

Farmers use AI-powered robots to help with harvesting of crops

There are three major ways which AI has been impacting the agriculture industry:

Agricultural Robots — Developing and programming autonomous robots to handle essential agricultural tasks such as harvesting crops at a higher volume and faster pace than human workers.
Crop and Soil Monitoring — Using computer vision and deep-learning algorithms to process data captured by drones and/or software-based technology to monitor crop and soil health.
Predictive Analytics — Developing machine learning models to track and predict various environmental impacts on crop yield such as weather changes.

With food becoming more scarce in this world, farmers will need to implement more technology to help with food production and the issues associated with farming such as their ever-growing lack of resources.

And that’s not all of the industries AI has been impacting…

If you enjoyed this article, please give it some claps! Follow me on LinkedIn or check out my personal website to learn more about other projects I’ve completed in the past!

Finally, if you have any constructive criticism or feedback, be sure to leave a comment below! I’d love to learn ways to improve my articles/writing style :)

Grade 12 high school student and BCI & VR developer! Feel free to visit www.ayleenfarnood.com to learn more about me :)