(Image source: Internet) Recently, the American quantum hardware company IonQ announced that it has achieved success in a collaborative project with its partner, the American General Electric Company Innovation Center (GE Research), in exploring the application of quantum computi

(Picture source: Internet)

Recently, the American quantum hardware company IonQ announced its cooperation with its partner, the American General Electric Company Innovation Center (GE Research), to explore the application of quantum computing to multi-variable distribution modeling in risk management. The project has achieved promising preliminary results.

IonQ revealed that based on the Quantum Circuit Born Machine (QCBM) framework of standardized, historical data indexes, they and GE Research used quantum circuits to effectively train and learn the correlation between three to four indexes. At the same time, found that in some cases predictions based on quantum frameworks were better than classical modeling methods. These results are strong confirmation that quantum copulas will enable smarter data-driven analysis and decision-making in commercial applications.

IonQ CEO and President Peter Chapman said: "IonQ is working with GE Research to advance the goals of applications that are now possible with quantum computing. While classical computing techniques face inefficiencies when modeling multiple variables with high accuracy, Our joint efforts have identified a new training strategy that optimizes quantum computing results even as the system scales, and has been tested on the IonQ Aria system. IonQ is excited to apply these new methods to handle excessive complexity. and unsolvable real-world scenarios."

While classic techniques of composing copulas using mathematical approximations are a good way to build multivariate risk models, they are limited when scaling. In response to this bottleneck, IonQ and GE Research successfully trained a model with up to four variables on IonQ's ion trap system by using data from four indexes that are easily accessible and the most representative stocks of changing market conditions. Quantum copula model.

By studying the historical basis structure between the four index returns over that time frame, the research team trained its model to understand the underlying dynamics. In addition, the newly proposed method includes optimization techniques that allow the model to scale by mitigating local minima and eliminating the vanishing gradient problem common in quantum machine learning practice.

Based on these improvements, they demonstrated a faster and more accurate way to perform multivariate analysis GE researchers will use it in the future to use newer and better methods to evaluate major manufacturing processes such as product design, factory operations and supply chain management) risks.

David Vernooy, senior executive at General Electric and head of digital technologies , said: "As we have seen from recent global supply chain fluctuations, the world needs more effective methods and tools to manage high levels of change and associated risks. The early results we have achieved in the financial use cases we are working on with IonQ show that quantum computing has great potential to better understand and reduce the risks associated with such variable scenarios."

original article. Link:

https://www.hpcwire.com/off-the-wire/ionq-and-ge-research-demonstrate-high-potential-of-quantum-computing-for-risk-aggregation/

Text: Jonathan Spencer Jones

Compiler: Li Mi

Editor: Mu Yi

Note: This article is compiled from "HPCWire" and does not represent the views of Quantum Outpost.