Quantum Feature Selection for Machine Learning

PROJECT BACKGROUND

Machine learning plays an increasingly important role in unlocking the ability to extract actionable insights from complex datasets across a broad range of fields. The collection of ever-increasing amounts of data brings new challenges in data management, preparation, model training, and interpretation.

With increasing data sets comes increasing complexity, especially in the fundamental building block of Machine Learning, feature selection. This program focuses on utilizing quantum hybrid approaches to optimize feature selection in model training and prediction.

D-Wave’s hybrid quantum computing service makes it possible to efficiently perform feature selection by constructing and solving a combinatorial optimization problem. Compared to classical alternatives like recursive feature elimination, this approach provides a “one-shot” solution and avoids the potentially costly cycle of iterative model training and incremental construction of the solution.

PROJECT DESCRIPTION

The D-Wave Launch™ program is designed to help enterprises go from problem discovery through production implementation. Under this project, D-Wave will work with you to develop a custom feature selection solution that utilizes D-Wave’s Leap™ hybrid quantum computing service. D-Wave’s professional services team members will work with you to design and implement a solution that is tailored to your use case. The solution approach will be informed by the characteristics of your data as well as the specific requirements and objectives for the feature selection solution and its role within your broader machine learning strategy.

Project activities will include the following:

• Collaborating to develop a deeper understanding of the relevant datasets, use cases, and objectives
• Identifying the data required to proceed with solution development and testing
• Evaluating and defining appropriate pre-processing and data encoding steps
• Developing a feature selection model to be used as the foundation of the solution
• Implementing the solution using D-Wave’s Ocean™ developer toolkit
• Testing the solution using data representative of the problem space As each business problem is unique, the exact duration necessary to produce the quantum feature selection solution is variable, however, average project duration is typically in the range of 3-6 months.

DELIVERABLES

During the course of this project, D-Wave will work with you to produce deliverables that demonstrate the value of a quantum hybrid application to solve your business problem. Types of deliverables include:

• Bi-weekly status reports
• Code required to prep data, build model, and submit to Leap
• Project summary report to include the following:
  • Problem definition
  • Problem formulation
  • Solution overview
  • Test results
  • Recommendations for future work
• Executive summary presentation to summarize the report details

LEAP CLOUD SERVICE

During the project, D-Wave will provide up to five (5) of your team members with access to D-Wave’s Leap quantum cloud service as needed for exploration and development related to the project.

TRAINING

To help your team get up to speed on the Leap service and to empower you to be able to create your own quantum hybrid applications, D-Wave will provide up to five (5) seats in D-Wave Learn™, our regularly scheduled, week-long, self-paced and instructor-led online training.

D-WAVE

D-Wave is a leader in the development and delivery of quantum computing systems, software, and services, the world’s first commercial supplier of quantum computers and the only company building both annealing quantum computers and gate-model quantum computers. Our mission is to unlock the power of quantum computing for business and society, today. We do this by delivering customer value with practical quantum applications for problems as diverse as logistics, artificial intelligence, materials sciences, drug discovery, scheduling, cybersecurity, fault detection, and financial modeling. D-Wave’s systems are being used by some of the world’s most advanced organizations, including NEC Corporation, Volkswagen, DENSO, Lockheed Martin, University of Southern California, Forschungszentrum Jülich and Los Alamos National Laboratory. With headquarters near Vancouver, Canada, D-Wave’s US operations are based in Palo Alto, CA. D-Wave has a blue-chip investor base that includes PSP Investments, Goldman Sachs, BDC Capital, NEC Corp., Aegis Group Partners, and In-Q-Tel.