Let’s talk about Artificial Intelligence
Artificial Intelligence (AI) is such a popular term that can be heard anywhere within decades.
But, what actually is AI?
Basically, it can be viewed as a combination of plentiful algorithms that can make decisions about data. Those algorithms are made with math and code that a lot of people studied in college. But the point is, most people only want to know what actually AI can do, not math and science.
Therefore, we don’t even have to know which algorithms are involved.
For a high-level understanding, AI makes predictions about something you care about or makes a system have the ability to adapt to the environment. For example, you might want to detect a tumor among patients’ X-ray images automatically. The tumor detection result is the outcome that you really care about. For now, we have strategic clarity about what you want to use AI for. Figure 1 is an example that shows a system which is used to detect anomalies within these X-ray images automatically.
Figure 1: patients’ X-ray images
When we need AI…
Once we have the target that needs AI, the next step is to find out whether you have data that correlates with what you want to predict. In this case, you might want lots of patients’ X-ray images. After gathering the image data, you want to detect the tumor within these X-ray images automatically. To perform this work, algorithms, especially machine-learning or deep learning algorithms, will be involved. That is, if you give these algorithms the outputs (detect a tumor!!!) and the inputs (X-ray images), they will find the rules that what kind of X-ray images has a tumor on it. Thus, the AI will be able to say: “This patient is healthy. No tumor detected.”
In a nutshell, AI is the field that aims at building. Build a system to match human performance, to match ideal rationality; or, build systems that can think and act. Considering the algorithms involved, we cannot talk about it without mentioning machine learning and deep learning, we will compare the differences between these three terms in our next article.
Editor: Chieh-Feng Cheng
Ph.D. in ECE, Georgia Tech
Technical Writer, inwinSTACK