The Machine Learning Revolution

What is Machine Learning?

Machine learning is the branch of artificial intelligence (AI) that deals with computers learning from data without explicit programming. As such, it is an ideal type of AI for extracting new meaningful insights from datasets. Machine learning has allowed advances in many other fields of computer science such as computer vision, language translation, and natural language processing. It holds the power to improve many existing consumer-facing technology products as well as to provide superior business intelligence to companies.

Where is it used?

We are entering a golden age of machine learning. It is already in use in several industries. If you use an app or platform whose analytics or recommendations you love, you’ve probably seen the power of machine learning firsthand. Facebook relies heavily on machine learning to tailor each user’s experience. It leverages the technology to ensure users see what is most relevant to them. They use it in ranking feeds, ads, and even search results..

An example of its cross-industry use is that the leading weather forecasting, flight price analysis, and stock analysis platforms all now make use of machine learning to provide insights for their users.

Hopper, a flight-price-prediction startup, began as a full text travel search system. Though well funded, their initial product did not gain wide user adoption. However, their blog post about analytics of when to book flights began driving most of their traffic. That allowed them to pivot into a focus on big data and machine learning flight price analysis. It was an instrumental turn that has led Hopper to raise over $60 million and propelled their growth to millions of users. Now, even their competitors are moving towards flight price change forecasts.

The merits of machine learning are no longer in question. At Raizlabs, we’ve begun our foray into several aspects of the machine learning revolution, from computer vision (with our own Smile-O-Meter, seen in the video below) to custom algorithms for our client projects. In fact, one of our developers has built a machine learning-powered daily stock analysis platform used by thousands of active traders around the world. The key questions now are: what are the most meaningful ways to make use of machine learning, and what can each business do to stay competitive in the machine learning age?

Our developer, Dan Murphy, demos the Smile-O-Meter, which uses machine learning.


Can it work for your team?

In terms of automation, the general rule of thumb is that any action which can be done in one second of cognitive thinking can be automated, explains Andrew Ng of Baidu Inc., in a recent Wall Street Journal article. However, many complex tasks can be sliced into several self-contained seconds of cognitive thinking. Thus, machine learning can be applied to a far wider range than most would assume. In fact, it has evolved from the automation of routine tasks towards the improvement of high level tasks with efficiency as high or higher than leading experts of those fields.

A recent article from Forbes magazine chronicled what they believe will be the top ten use cases for machine learning. The use cases from its list that are applicable to any company are:

  • Marketing Personalization
  • Recommendations
  • Natural Language Processing

In the case of marketing personalization, a scenario may be an increase in brand loyalty because it could make the consumer feel that the company marketing to them is better tailored to their needs.

Likewise for recommendations. However, recommendations don’t need to be limited to client-facing applications. They can also be a powerful differentiator inside a business. Appropriate machine learning of a business’ data for recommendations can reduce costs by enabling more informed resource allocation and even increase revenue by providing key insights for business development.

Machine learning provides a wealth of new, meaningful information about data. Natural language processing enables people to have easy access to the information they need through voice and text chat bots. This makes the everyday use of machine learning more practical and seamless. It allows people to interact with novel data naturally and at their convenience.

The great opportunities emerging from machine learning are not without pitfalls. These incredible new technologies have the potential to disrupt many existing industries. Plus, meaningful insights for business require the combination of quality algorithms and data, which a business may prefer to keep private.

Though there are available platforms that may perform machine learning on data, it is important to be mindful that any data a business prefers to keep private does not fall in the wrong hands. The best uses of machine learning may not be from the use of generic algorithms but rather from the crafting of new machine learning technology to extract insights uniquely useful for a particular business or product.

Unlike prior technological revolutions, this is one where each company may need to lead with its own creative use of machine learning to improve its business’ competitiveness in the market. As this technology becomes prevalent in more and more applications, consumers and end users will come to expect higher quality experiences enabled by machine learning.

Machine learning may very well be the technology that will have the biggest impact in the following years, as there is immense value to be created from its use. If you’re interested in exploring machine learning and have an idea for a project, the team at Raizlabs can help. Get in touch!