An enhanced approach for predicting air pollution using quantum support vector machine.

Area of Science:

• Quantum Computing
• Machine Learning
• Environmental Science

Background:

• Classical Support Vector Machines (SVM) face limitations with complex datasets.
• Quantum machine learning leverages quantum mechanics for enhanced computation.
• Air quality prediction is crucial for public health and environmental monitoring.

Purpose of the Study:

• To compare the accuracy and execution time of classical SVM and quantum SVM for air quality prediction.
• To introduce and evaluate a novel method for selecting optimal quantum feature maps.
• To demonstrate the potential of quantum-enhanced feature mapping to overcome classical SVM constraints.

Main Methods:

• Utilized conventional SVM to identify an optimal feature map and benchmark dataset for air quality prediction.
• Implemented and compared classical SVM and quantum SVM algorithms on a shared dataset.
• Conducted experiments using IBM's quantum computer cloud for performance benchmarking.

Main Results:

• Quantum SVM achieved higher accuracy (97% and 94%) than classical SVM (91% and 87%) in air quality prediction.
• The selection of appropriate quantum feature maps significantly impacts classification performance.
• Quantum-enhanced feature mapping demonstrated superior effectiveness compared to classical approaches.

Conclusions:

• Quantum SVM presents a more accurate and effective method for air quality assessment.
• The study validates the advantage of quantum computing for complex machine learning tasks.
• Optimized quantum feature mapping is key to unlocking the full potential of quantum SVM.