What is Quantum Machine Learning?
Most are or are becoming familiar with artificial intelligence (AI), machine learning (ML), and quantum computing, but among these are fast developing a demanding field in quantum machine learning?
Because some machine learning algorithms may be too taxing for classical computers, quantum machine learning (QML) would potentially allow scientists to take a classical machine learning algorithm and translate it into a quantum circuit so it can be run efficiently on a quantum computer, according to The Quantum Daily.
The development and potential capabilities of quantum computing will likely accelerate advances in current computing paradigms, including AI. Much more than simply a conceptual combination of technologies, quantum machine learning is an emerging field that will develop quantum algorithms to perform advanced machine learning tasks.
While IBM, Intel, Google, and other major technology corporations have recently made significant advancements in quantum computing, there are still many hurdles (and not all of them are technical) before the technology will be a practical resource for businesses.
In an August 2020 interview with Military & Aerospace Electronics, Tatjana Curcic, program manager for Optimization with Noisy Intermediate-Scale Quantum devices (ONISQ) of the U.S. Defense Advanced Research Projects Agency in Arlington, VA, said, “Quantum machine learning is a very active research area, but is quite new,” further explaining, “the interface between classical data, which AI is primarily involved with, and quantum computing is still a technical challenge.”
Google’s TensorFlow Quantum
One of the challenges of harnessing the powers of quantum computers is to reimagine existing machine learning models to work on quantum architectures. To facilitate that process Google opened sourced TensorFlow Quantum (TFQ), which provides quantum algorithm research and ML application research with a Python framework to build quantum machine learning models by leveraging Google’s quantum computing frameworks.
Using quantum data and by building hybrid quantum-classical models, TFQ is designed to “provide tools to interleave quantum algorithms and logic design in Cirq with TensorFlow,” according to TensorFlow.org.
IBM’s Quantum Challenge
Designed to help classic software programmers become quantum ready developers, IBM’s Quantum Challenge is an annual multi-day event in which software developers, researchers, and business users learn more about how quantum programming works through a series of exercises.
Approximately 1,745 people from 45 countries participated in the 2020 IBM Quantum Challenge and the participants total use of the 18 IBM Quantum systems on the IBM Cloud reportedly exceeded 1 billion circuits a day.”
In an interview with TechRepublic, Abe Asfaw, Global Lead, Quantum Education and Open Science at IBM, said that “a quantum computer is going to work in union with a classical computer. And at the end of the day, what we hope to achieve from this process is that everyone is equipped to be able to program a quantum computer.” To that end, IBM quantum learning materials, including a textbook and video series, for their open software, Qiskit, is free and accessible at qiskit.org/education.
You can also explore snippets of code, write your code, and test it yourself on a real quantum computer at quantum-computing.ibm.com.
Randomness Breakthrough and Quantum Computing
In September 2020, InsideHPC reported Cambridge Quantum Computing (CQC) and IBM announced the first cloud-based quantum random number generator (QRNG). The history-making application generates entropy (true maximal randomness) that can be verified and therefore, certified as truly quantum.
The breakthrough is significant on several levels, not the least of which is because randomness is an omnipresent and essential element in the majority of digital interactions. It is also key to data encryption, and therefore, cybersecurity. Finance, pharmaceutical, healthcare, military, telecommunications, and gaming are among the many sectors that also depend upon secure data and communications.
The beta-certified QRNG is a milestone in quantum computing as well as in commercializing quantum computing through cloud delivery to provide real-world application of the technology.
CQC was a founding member of the IBM Q Network which was created in 2018, and in January 2020, IBM invested in CQC. CQC also works with other members of the IBM Q Network on chemistry, optimization, finance, and quantum machine learning and natural language processing.
Educational Background for Quantum Computing
While PhDs in physics and electrical engineering have traditionally dominated quantum computing, IBM and Google have been aiming their quantum educational efforts toward undergraduate college level. While programming experience and knowledge of linear algebra are basic requirements to understand and explore quantum computing, individuals in other STEM (Science, Technology, Engineering, and Math) fields have an interest in developing quantum algorithms.
Data scientists may also be motivated to learn more quantum computing in order to study quantum machine learning.
Quantum Computing and AI
Developing and deploying AI and quantum computing is a global race and in the same interview with Military & Aerospace Electronics, Tammy Carter said, “AI is now a technology in deployment. Machine learning is pretty much in use worldwide. We’re in a migration of figuring out how to use it with the systems we have.”
Baidu Research predicts that the application of quantum algorithms artificial intelligence will experience great development, particularly as an increasing number of industry giants continue to invest in research and development (R&D) resources.
Machine Learning and Quantum Machine Learning
Now that we know more about quantum computing, how does machine learning fit in to make quantum machine learning? First, you need to know the basics of machine learning.
There are three types of machine learning: supervised learning (classification and regression), unsupervised learning (clustering and dimensionality reduction), and reinforcement (semi-supervised learning).
In a milestone discovery, IBM and MIT revealed the first experimental proof that the theory of combining quantum computing and machine learning could become reality. Published in Nature in March 2019, the study used a two-qubit quantum computing system to demonstrate that quantum computers could bolster classification supervised learning using a lab-generated dataset.
The Future of Quantum Machine Learning
The future is here but still has a long way to go before there are defined career or educational paths to quantum machine learning but with companies like Google and IBM providing open-source software, access to their quantum computers, and free educational tools, the technology will continue to advance.
In an exclusive interview with MIT Technology Review Google CEO Sundar Pichai said, “We think AI can accelerate quantum computing and quantum computing can accelerate AI. And collectively, we think it’s what we would need to, down the line, solve some of the most intractable problems we face, like climate change.”
Pichai stressed the importance of helping the public understand that quantum machine learning is in very early stages of development and he believes that most of the problems in the world will continue to be solved with classical computing.