In our last chapter (Chapter 1), we reviewed what Artificial Intelligence was by providing an overview. Specifically, the following topics were covered:
■ An introduction to Cybersecurity;
■ The various aspects of Cybersecurity;
■ A chronological timeline into the evolution of Cybersecurity;
■ An introduction to Artificial Intelligence;
■ A definition of Artificial Intelligence;
■ The various components of Artificial Intelligence and their technical definitions (this includes the likes of Machine Learning, Computer Vision, and Neural Networks);
■ An overview into the book;
■ Tlie history of Artificial Intelligence;
■ The importance of data and its role with Artificial Intelligence systems and applications;
■ The applications of Artificial Intelligence.
In this chapter, we examine the very first subcomponent of Artificial Intelligence, which is that of Machine Learning, also known as “ML” for short. We will do a deep dive first in the theoretical aspects of Machine Learning, and then this will be followed by the various applications, just like in the last chapter. But before we start getting into all of the theoretical aspects of Machine Learning, we will first provide a high level overview of what it is all about.
The High Level Overview
Although Machine Learning has been around for a long time (some estimates have it as long as a couple of decades), there are a number of key applications in which Machine Learning is used. Some examples of these are as follows:
1) Predictive Maintenance:
This kind of application is typically used in supply chain, manufacturing, distribution, and logistics sectors. For example, this is where the concept of Quality Control comes into key play. In manufacturing, you want to be able to predict how many batches of products that are going to be produced could actually become defective. Obviously, you want this number to be as low as possible. Theoretically, you do not want any type or kind of product to be defective, but in the real world, this is almost impossible to achieve. With Machine Learning, you can set up the different permutations in both the mathematical and statistical algorithms with different permutations as to what is deemed to be a defective product or not.
2) Employee Recruiting:
There is one common denominator in the recruitment industry, and that is the plethora of resumes that recruiters from all kinds of industries get. Consider some of these statistics:
■ Just recently, Career Builder, one of the most widely used job search portals reported:
- * 2.3 million jobs were posted;
- * 680 unique profiles of job seekers were collected;
- * 310 million resumes were collected;
- * 2.5 million background checks were conducted with the Career Builder platform.
- (SOURCE: 1).
Just imagine how long it would take a team of recruiters to have to go through all of the above. But with Machine Learning, it can all be done in a matter of minutes, by examining it for certain keywords in order to find the desired candidates. Also, rather than having the recruiter post each and every job entry manually onto Career Builder, the appropriate Machine Learning tool can be used to completely automate this process, thus freeing up the time of the recruiter to interview with the right candidates for the job.
3) Customer Experience:
In the American society of today, we want to have everything right here and right now, at the snap of a finger. Not only that, but on top of this we also expect to have impeccable customer service delivered at the same time. And when none of this happens, well, we have the luxury to go to a competitor to see if they can do any better. In this regard, many businesses and corporations have started to make use of Virtual Agents. These are the little chat boxes typically found on the lower right part of your web browser. With this, you can actually communicate with somebody in order to get your questions answered or shopping issues resolved. The nice thing about these is that they are also on demand, on a 24/7/365 basis. However, in order to provide a seamless experience to the customer or prospect, many business entities are now making use of what are known as “Chat Bots.” These are a much more sophisticated version of the Virtual Agent because they make use of Machine Learning algorithms. By doing this, the Chat Bot can find much more specific answers to your queries by conducting more “intelligent” searches in the information repositories of the business or corporation. Also, many call centers are making use of Machine Learning as well. In this particular fashion, when a customer calls in, their call history, profile, and entire conversations are pulled up in a matter of seconds for the call center agent, so that they can much easier anticipate your questions and provide you with the best level of service possible.
In this market segment, there is one thing that all people, especially the traders, want to do, and that is to have the ability to predict the financial markets, as well as what they will do in the future, so that they can hedge their bets and make profitable trades. Although this can be done via a manual process, it can be a very laborious and time-consuming process to achieve. Of course, we all know that the markets can move in a matter of mere seconds with uncertain volatility, as we have seen recently with the Coronavirus. In fact, exactly timing and predicting the financial markets with 100 percent accuracy is an almost impossible feat to accomplish. But this is where the role of Machine Learning can come into play. For example, it can take all of the data that is fed into it, and within a matter of seconds make more accurate predictions as to what the market could potentially do, giving the traders valuable time to make the split-second decisions that are needed to produce quality trades. This is especially useful for what is known as “Intra Day Trading,” where the financial traders try to time the market as they are open on a minute-by-minute basis.