Quantum Computing refers to the use of quantum mechanical phenomena to make computations. This field is making big strides in the last decade because it can actually help us solve some of the most challenging problems in the realm of computer science, particularly in machine learning and security. Machine learning is all about building better models of the world to make more accurate predictions and security is about safeguarding the things we have built. For example, if we want the machines to see things better, we need better models of how we process visual data. If we want to understand currency fluctuations, we need better models of how they change over time. If we want to create effective environmental policies, we need better models of what’s happening to our climate. So how can we use quantum computing to do these things?
Why machine learning in particular?
To start off, machine learning is really difficult. It’s what mathematicians refer to as “NP-hard”. What it means is that once we have a solution, we can verify whether or not it’s right very quickly. But locating that solution in the first place is really hard. Coming back to machine learning, it’s difficult because building a good model is not an exact science. It’s actually a creative act! For example, let’s say we want to architect a house. You’re balancing lots of constraints like budget, usage requirements, space limitations, etc. At the same time, we are trying to create the most beautiful house we can. A creative architect will find a great solution. Mathematically speaking, the architect is solving a constrained optimization problem. We can think of creativity as the ability to come up with a good solution given an objective and the constraints.
What’s wrong with today’s computers?
The problem with our computers today is that they aren’t well suited to these types of creative problems. These computers belong to the realm of classical computing. Solving creative problems can be imagined as trying to find the lowest point on a surface covered in hills and valleys. Classical computing might use what’s called “gradient descent”. What happens here is that we start at a random spot on the surface, look around for a lower spot to walk down to, and repeat until you can’t walk downhill anymore. This works great if the starting spot is chosen right. But more often than not, the computer gets stuck in a “local minimum”, i.e. a valley that isn’t the very lowest point on the surface.
How does quantum computing solve this issue?
We saw that if we use classical computing techniques, we will not be able to reach the global optimum in reasonable time. That’s where quantum computing comes in! It lets you cheat a little, giving you some chance to “tunnel” through a ridge to see if there’s a lower valley hidden beyond it. This gives you a much better shot at finding the true lowest point, the global optimum. There are already some quantum machine learning algorithms that produce very compact, efficient recognizers. This is particularly useful when you’re short on power, like on a mobile device.
Another good thing about using quantum computing is that it can handle highly polluted training data. In machine learning, we use data to train our algorithms. Once the training is over, the algorithms utilize that knowledge to extract information about unknown data. In the real world, where a high percentage of the examples are mislabeled, quantum computing is very useful. Another interesting thing is that we get the best results not with pure quantum computing, but by mixing quantum and classical computing.
Building quantum computing hardware is a very expensive process. Google has taken a head start and they have created a Quantum AI Lab in partnership with NASA and D-Wave. The hope is that it would help researchers construct more efficient and more accurate models for computer vision, speech recognition, web search, neuroscience, protein folding, etc. Quantum machine learning may provide the most creative problem-solving process under the known laws of physics. Let’s see what comes out of that lab!
————————————————————————————————-
Perpetual Enigma 