Feature Selection for Recommender Systems with Quantum Computing

Area of Science:

• Quantum Computing
• Recommender Systems
• Optimization Problems

Background:

• Quantum computing's potential has been largely theoretical due to hardware limitations.
• Small, functional quantum computers are now accessible, enabling practical applications.
• Quantum annealing is a paradigm suited for solving NP-hard optimization problems.

Purpose of the Study:

• To design a hybrid feature selection algorithm for recommender systems.
• To leverage user interaction data and domain knowledge.
• To solve feature selection as an optimization problem on a quantum computer.

Main Methods:

• Developed a hybrid feature selection algorithm for recommender systems.
• Formulated feature selection as an optimization problem.
• Utilized a D-Wave quantum computer for solving the optimization problem.

Main Results:

• The proposed hybrid algorithm effectively selects a limited set of important features.
• Demonstrated the practical applicability of quantum computers in applied science.
• Showcased the potential of quantum annealing for recommender system optimization.

Conclusions:

• Quantum computers are maturing for real-world scientific applications.
• Hybrid quantum-classical approaches are viable for complex problems like feature selection.
• Quantum annealing provides an accessible method to explore quantum computing's capabilities.