Quantum Annealing is a way to use quantum physics's intrinsic effects to help solve a particular type of problem. Probability Sampling and Optimization are the types of problems that Quantum Annealing can help solve. Optimization means you try to search for the best configuration to find a solution from many different possible combinations.
For example, suppose you try to build a house. You have only a fixed budget, and there are many options and combinations of choices available to choose from. You want to find the best configuration for your fixed budget. Do you want to spend your entire budget on a great house or a home that is not so good? Try to use this example for the project that is scaled up. The base logic remains the same. An optimization problem will help you find the best configuration.
The reason you can solve physics to solve an optimization problem is that
you can frame the issue as an Energy Minimization problem. In physics, everything is trying to find its minimum energy state. In thermodynamics, hot things cool down.
Quantum annealing is using quantum physics to find the minimum energy state of something. However, Sampling problems are related to optimization problems with a difference in approach.
Instead of focusing on the minimum energy state, you sample from any low energy states that try to characterize the shape of your energy landscape.
This is extremely useful in fields such as machine learning where you are trying to build a probabilistic representation of the problem in the real world. These samples will provide information about your current and future models and use them to improve your models over an extended time.
In machine learning, sampling from optimization problems is challenging and extremely time-consuming to perform using the classical computers—alternative techniques employing Quantum Annealing approach.