SURREY, British Columbia, November 12, 2024 -- Today, the Quantum Algorithms Institute (QAI) announced a partnership with Canadian companies, AbaQus and InvestDEFY Technologies, to solve common challenges in training machine learning models for financial applications using D-Wave‘s quantum annealing systems.

The partnership aims to tackle complex problems in financial forecasting and predictive analytics such as non-stationary data, model overfitting, and generalization. Progress in these aspects make financial machine learning more streamlined by eliminating unnecessary features in financial datasets, improving the speed and accuracy of financial forecasts.

The collaboration has successfully identified key feature analysis issues in machine learning model designs and reformulated them for execution on quantum computers. Test runs on D-Wave’s quantum systems have already demonstrated useful results:

Optimized Feature Subset Selection: Preliminary tests indicate that quantum annealing can improve the identification of data or feature sets for model performance compared to some classical methods.

Reduction in Computational Time: The approach shows promise in reducing the time taken to evaluate large data sets, though precise metrics are still being established.

“The financial industry deals with complex datasets, countless variables and market shifts, making it challenging to build more effective predictive models,” said Louise Turner, Chief Executive Officer at QAI. “Our collaboration with D-Wave uses quantum annealing to optimize ‘feature selection’—identifying the most relevant data points to inform the predictive model more efficiently than existing solutions performed on classical computers.”

This ongoing partnership aims to showcase the potential of quantum computing in real-world financial modeling, leading to breakthroughs in how well financial prediction models work when data patterns change over time.

“This collaboration allows us to bring quantum computing’s power to an industry where every decision counts. By refining feature selection, we’re helping financial models become both faster and more accurate,” David Isaac, Co-founder and CEO of AbaQus Computing.

“We are very excited to explore and incorporate quantum methods within our data science platform with the objective of improving predictive model performance,” noted James Niosi, CEO of InvestDEFY.