Where can you use quantum computing in data analysis?
Aleksandar Lazarevic, VP Advanced Analytics & Data Engineering, Stanley Black & Decker and Aaron McClendon, Data Scientist and Practice Leader, Aimpoint Digital
Aaron McClendon is a Data Scientist and Practice Leader at Aimpoint Digital, specializing in advanced applications of machine learning and artificial intelligence within enterprises facing data science challenges. His team focuses on time series analysis, natural language processing, deep learning and AI in a research context to include quantum machine learning and potential applications of quantum computing within companies.
Aleksandar Lazarevic is vice president of advanced analytics and data engineering at Stanley Black & Decker. Aleks leads the company’s efforts to drive cost savings through advanced big data analytics. Aleks is responsible for delivering a scalable enterprise-scale data lake/data warehouse platform and leveraging machine learning/AI tools to solve various business problems. Previously, Aleks worked at Aetna, where he was responsible for the global analytics solution for healthcare fraud, waste and abuse detection. Additionally, he has extensive experience in applying analytics in various industries ranging from banking, credit and insurance to smart manufacturing and IT security. He is also a frequent speaker at national data science and analytics conferences.
Quantum computing is a disruptive and revolutionary new type of technology that will soon have major impacts on the way we do business. From the printing press to the locomotive, each revolution has dynamically forced our hand to change the current economic operating model. During each revolution, the winds of success blow benevolently on those who are quick to adapt and learn and drive away those who retard and retard.
Although still in its infancy, quantum computing has the potential to impact such crucial operations as logistics or large-scale traffic control, enabling, in some cases, exponentially faster computing power. and billions of dollars in additional revenue.
A quantum computer works using quantum bits, or qubits. Unlike a typical computer bit which can only represent a 0 or a 1, a qubit can exist in a state of superposition, meaning it can be a bit 1 and a bit 0 at the same time. Only by “looking” at it does it become all 1s or all 0s. This may allow a single qubit to represent multiple things at the same time, rather than the given deterministic state in a classical computer. This is where the power of quantum computers lies.
There are many applications where this becomes practical and we must consider several possibilities at once. Consider a classical traveling salesman problem where finding an optimal solution that minimizes travel time is computationally very expensive. A typical computer has to consider all possible routes, which, if you remember your high school math lesson on permutations, becomes an incredibly large number as we increase the number of sites it has to visit. Quantum computers are believed to be able to solve these types of problems exponentially faster because they can consider multiple paths at once. Although the above example may seem contrived, it applies directly to day-to-day issues such as logistics, travel routes, supply chain, and inventory optimization. Imagine the value generated by knowing exactly the optimal route to move shipments between factories and stores. Until now, finding the perfect route has been a pipe dream.
Quantum computing is already being researched and applied in many industries. For example, in Barcelona, Spain, a city buzzing with taxis, subways and personal cars, Volkswagen worked with a quantum computer from D-Wave (a leading company in the development of quantum computers), to develop solutions for quantum computing to minimize traffic allowing companies to best dispatch taxis and cars. This allowed transport companies to avoid long journeys with empty seats, avoided long waiting times for taxis and benefited the whole population by reducing traffic jams. In another example, IT giant Microsoft has partnered with the Dubai Electricity and Water Authority to optimize electricity distribution. Currently, finding an optimal solution for the ideal balance of electrical resources is a difficult task for modern computers. However, quantum algorithms are currently being used to test multiple scenarios at the same time, improving power grid distribution exponentially faster.
In addition to the aforementioned examples, there are many other analytics use cases where quantum computing could find the solution much faster. Natural language processing (NLP) is one such application where computationally intensive topic modeling could be used to generate value in new product development. When a new product is launched, it often takes a year or more to understand the public reaction to the article. Instead of waiting for sentiment to be readily available, we could search the web for all relevant product information and reviews from multiple places, including online retail stores, online shopping reviews , professional review sites, social media, news and financial releases. This might allow us to identify common topics/themes that appear in these documents. For example, we can see which new product features users liked and which features were generally avoided. Quantum computing has recently begun to venture into this field in the form of quantum latent semantic analysis and quantum NLP. It has been shown that quantum subject analysis can outperform classical computer algorithms in some situations and could potentially allow larger amounts of data to be processed in less time than a classical computation would take.
Another very large area where quantum computing can have a significant impact is freight logistics. Often large companies must move products, both components and finished goods, between locations using a variety of different modes of transportation. Moving components through factories, shipping them across oceans, and loading ships and land transport with the right items and giving them the routes that best minimize time and cost is an area of intensive research in business. If a business knew the optimal routes to move inventory, it would enable faster response times to meet customer demand, thereby boosting sales while reducing inventory costs. Various companies have already started experimenting with quantum computers to secure advantageous freight routes. Quantum optimization programs are not only more computationally efficient, but also produce better results. The technology is still a few years behind large-scale implementation, but the first companies to develop the techniques and prove their superiority will certainly gain an edge over their competitors.
We are truly on the precipice of a new computing era, but there are still many hardware and software challenges before we can successfully apply quantum computers. First, they must be cooled to almost absolute zero. Second, they require a myriad of cutting-edge technology, and building them can easily run into the millions. Third, using a quantum computer requires manually stringing together qubits, about the lowest possible level of computer programming. This means that flashy machine learning packages and even standard programming commands taken for granted in classical computers do not exist in the quantum world.
Commercial applications of quantum computers are still being discovered and tech giants such as Google, IBM and Microsoft have even opened up their machines to the public for free use as people test the waters and begin to understand the uses and applications. Although quantum computers are unlikely to ever replace classical computers, a future in which the two work side by side is almost inevitable. Establishing quantum algorithm research divisions to research and develop new applications will propel companies into the future, putting them at the forefront of technology development and light years ahead of the competition.