What Is Random Walk?

You've probably heard of the random walk. It's a mathematical construct used in computer science and machine learning, and now it's all over your newsfeed. The concept is simple: it's a "stochastic" process for modeling how an intelligent or digital "rational actor" takes incremental steps forward. You can think of it like this: The random walk is like a drunk friend trying to get from point A to point B but getting lost on the way there. The random walk is a mathematical construct in computer science and machine learning. It is described as an "imaginary" process because it applies random unsteady. The random walk tracks incremental steps by a particular modeled intelligence or digital "rational actor." The arbitrary walk concept was first developed by Karl Pearson in 1905, who sought to create a model for how stock prices fluctuate over time. Pearson's approach was to take an average of past values and then add a random variable to that average, which might result in an increase or decrease from the previous value. This approach would continue until it reached a steady state where no further changes occurred. The same basic idea can be used when modeling human behavior, for example, by creating a model for how people make decisions based on prior experiences and then adding some randomness to those decisions based on the current environment or situation. In addition to helping model human behavior, random walks have also been helpful in other areas, such as image processing, robotics, and machine learning. The random walk is a perfect example of rational actor choices in machine learning. In game theory, the random walk represents an approach to predicting outcomes in machine learning systems. Studying some of the classic games studied by mathematicians in machine learning research shows how this approach can benefit a broad spectrum of projects.