How We Teach Machines to Learn: Tree Boosting and Adversarial Networks
No matter how you prefer to learn, you have probably heard someone mention something about your particular “learning style.” This seemingly simple phrase implies that there are many different ways to learn, and people can be better or worse at certain learning methods depending on personal preferences and circumstances. As it is with people, so we tend to make it for computers.
Machine learning is moving forward at a fast rate thanks to researchers and programmers figuring out how to optimize the ways through which machines learn new information. Two methods in particular, called tree boosting and adversarial networks, work in tandem to produce encouraging results. Read on to learn more about these two deep learning methods.
Making Well-informed Decisions: The Basics of Tree Boosting
Many of us have thought about it: one minor decision we make locking us into some indeterminate future that we seemingly cannot control, and the way we accept that there may not have been a choice at all. Then we start to think about all the choices we make on a daily basis, and these choices pile up. A basic version of this branching decision-making is the principle underlying tree boosting, otherwise known as gradient boosting.
To put it simply, to induce machines to learn, the computers make decisions, which are sometimes as simple as “correct” or “false” scenarios. When a computer makes the “wrong” choice, giving the “incorrect” answer, the result is called a loss. To minimize losses, engineers came up with a technique called tree boosting. Picture a tree being the initial problem, and the two possible answers branching from the tree. To minimize the number of losses the computer experiences in its learning, programmers created more trees that could move from the initial question and answer to give the machine another chance to get it right, thus correcting the loss. Eventually, the trees’ branches will lead to a correct answer; this is the basic gist of tree boosting.
Competing for Accuracy: Adversarial Networks Usher in Correct Identification
For computers to learn the best ways to utilize information, you must train them to identify and classify images and information correctly. A new method called adversarial networks streamlines and improves these classification processes by pitting two aspects of a computer against themselves; these programs are called the generator and the discriminator, respectively. The generator’s job is to generate images that look real, while the discriminator must determine which of those images is a fake.
Ideally, this game played between the generator and the discriminator increases the computer’s intelligence quotient, as one program must make better and better “fakes” while the other must get better and better at working out the “fakes.” Think of it as the way a friendly competition brings out the best in each competitor; the competing aspects of the computer improve each other until the computer is more intelligent.
It is clear that computers are improving much faster than we ever thought possible, and new methods of teaching AI to think, such as tree boosting and adversarial networks, bring us and our machines closer to realizing true artificial intelligence.
Visit www.PROPRIUS.com for more information on how to improve your team and career. PROPRIUS is an Artificial Intelligence Industry recruiting firm dedicated to projecting organizations to the next level.
If you are ready to accelerate your team, then schedule a 10-minute discovery call at https://PROPRIUS.as.me/Discovery. We have a dedicated search process designed to locate, connect with, and deliver the most talented candidates.
If you are looking to propel your career, then schedule a 30-minute intake call at https://PROPRIUS.as.me/Intake. We identify the top Engineers in the Artificial Intelligence Industry that generate results, create opportunity and inspire others to perform their best work.
Joshua Crawford | Managing Director | PROPRIUS