What is Machine Learning? And How Are Companies Using It?

Some Machine Learning History.

If machine learning is the future…..the future is here. But to be fair, the term machine learning has been around since 1950’s and even though it has evolved over the years, machine learning has kept the core foundation of its meaning. In 1952, Arthur Samuel, an expert at IBM in computer gaming and artificial intelligence created the first learning program in the form of the game checkers. The more it was played the better the program improved at the game. It was able to learn strategies and incorporate moves to win into its program. After a few years of improving this program and introducing new applications for it, He and IBM coined the term “Machine Learning” in 1959.

Let’s fast forward a few years…

In the 1990’s machine learning makes a conversion from a knowledge based system to a data driven approach. They began to create programs for computers to be able to examine large amounts of data to make predictions and to draw conclusions from the results. This was a game changer, and computers actually started to learn from the information or data it was given.

Jumping a few more years ahead.

2004…the year of Facebook. Back then it was called “The Facebook”, and opened a plethora of applications for how machine learning could be used. 10 years later in 2014 Facebook designs a software called DeepFace that uses an algorithm that allows facebook to verify or recognize individuals in photos equivalent to the ability of a human. And that’s just the tip of the iceberg.

In 2015 Amazon Web Services (AWS), which is Amazon’s cloud computing platform, launches Machine Learning as a service.

That’s a long way from getting owned by a computer game of checkers.

So let’s break down what machine learning is exactly.

Machine learning is in the territory of computer science that enables computers the capacity to learn without being explicitly programmed. It is what some consider a classification of artificial intelligence which can carry out learned tasks based on its data analysis. Learning is a profoundly important part of what makes us human. However, if we able to build AI that can carry out tasks with human-like intelligence, we can in turn make machines that can learn for themselves, based on their past experiences and information that we give it. Exploring the study and construction of algorithms, machine learning can learn and make data-driven predictions and recommendation based on habits of the user.

These days machine learning is everywhere and even though the term has been around short of 60 years,  it has just recently started to have a substantial impact on large companies over the last decade. We can only expect it keep gaining a whole lot more traction in the market as we see new consumer interactive apps being built. It seems like more and more companies are catching wind and starting to adapt machine learning into a large part of their business. Machine learning is able to take large amounts of data and give us very accurate insites, make predictions based on data analytics, pattern recognitions, and even make product recommendations based on user habits. And that’s just a broad overview of what machine learning can do.

Some will argue that machine learning is not artificial intelligence. I would have to agree. However machine learning is considered a subcategory of AI based on the fact that it can learn without assistance and that in itself is a human like characteristic.

Big companies that use machine learning as a large part of their business strategy.

Amazon

Amazon has been investing in artificial intelligence for over 20 years to create an unmatched customer experience. But not only have they used machine learning as a strategy to implement product recommendation, as well as other aspects of machine learning. They release AWS or Amazon Web Services as a product for other businesses to be able to use the same type of strategies they use on their eCommerce store.

Amazon’s mission for machine learning is simple. They want to use their findings and share their machine learning capabilities as a fully managed service and implement it into every business, developer, and data scientist they can. This could be why Amazon is one of the top leaders in this field.

Netflix

Here is another large company that uses machine learning as a huge part of their strategy to keep their customers engaged. Netflix has over 100 million users including myself. I can’t get enough sometimes. Their user experience is not like any other TV/movie streaming system out there, which is most likely why it’s the most popular. Netflix says that 80% of TV shows its customers watch are discovered through its machine learning platform and recommendation system. I mean will even see when scrolling through lists of movies and TV shows categories named “Because you watched….” or “Top picks for….”. Netflix is not subtle about their recommendation system to say the least, and that’s a strategy that works great for them.

Netflix uses machine learning and algorithms to help break down its viewers behaviours on its platform. To go even further, it’s not just making recommendations to you based on just your habits but also a cluster of viewers habits. To make these kind of predictions for its customers, Netflix uses 3 elements; Netflix members, taggers who understand the content, and finally machine learning algorithms that put all the data together.

Two very large, and different companies with two very different strategies for machine learning as it applies to their business.

Conclusion

As we can see machine learning is a method that can use data differently for analysis that can automate analytical model building in ways that can be customized based on your business strategy. It is not artificial intelligence but rather a branch or sub-set of artificial intelligence. Machine learning is very much based on the idea that machines are able to learn and adapt unsupervised through experience. Machine learning is becoming the foundation of how companies run their business, and machine learning platforms are even being sold as a stand alone service to improve the functions of your company. If you are looking to implement a machine learning platform for your business, we can help with that here at Trifin Labs Inc. We have experts that have an immense amount of knowledge on machine learning. If you are looking to speak with one of our expert, contact us and get a FREE 15 min consultation with one of our experts.