1323-UR Evaluating Quantum Machine Learning for Civil and Environmental Applications
Location
https://campus.kennesaw.edu/colleges-departments/ccse/events/computing-showcase/fa26-cday-program.php
Document Type
Event
Start Date
20-11-2026 2:00 PM
Description
We compare seven quantum machine learning families with classical models across eight civil and environmental datasets covering concrete strength, crack detection, air quality, drought, land cover, transportation and building energy. Experiments use classical simulation and each module's documented evaluation protocol. Quantum methods include neural circuits, kernels, decision trees, reservoirs, Bayesian classifiers, annealing and principal component analysis. At 17 fitted coefficients, the quantum concrete-strength kernel reduces test RMSE from 8.192 to 7.753 MPa against an RBF kernel. Boosted trees achieve 4.847 MPa. Tree and Bayesian circuits match classical counterparts, while reservoir and quantum PCA comparisons show similar performance. The results identify model-size tradeoffs under task-specific conditions.
1323-UR Evaluating Quantum Machine Learning for Civil and Environmental Applications
https://campus.kennesaw.edu/colleges-departments/ccse/events/computing-showcase/fa26-cday-program.php
We compare seven quantum machine learning families with classical models across eight civil and environmental datasets covering concrete strength, crack detection, air quality, drought, land cover, transportation and building energy. Experiments use classical simulation and each module's documented evaluation protocol. Quantum methods include neural circuits, kernels, decision trees, reservoirs, Bayesian classifiers, annealing and principal component analysis. At 17 fitted coefficients, the quantum concrete-strength kernel reduces test RMSE from 8.192 to 7.753 MPa against an RBF kernel. Boosted trees achieve 4.847 MPa. Tree and Bayesian circuits match classical counterparts, while reservoir and quantum PCA comparisons show similar performance. The results identify model-size tradeoffs under task-specific conditions.