How twricks/yquantum-2026-vaoa-vs-dqi works
This repository compares classical ILP, DQI, and QAOA methods for P&C insurance package optimization. The main workflow loads insurance instance data, solves a classical baseline, runs quantum approaches package by package, and records results for comparison. The README documents the DQI parity-check and decoding stages and the QAOA variational loop; sampled excerpts confirm key model, execution, and evaluation responsibilities. Some internal wiring is documented rather than fully visible in the partial source excerpts, so uncertain links are omitted.
Main entry points: classical.py ILP Solver · insurance_model.py Bundling Model · dqi.py Circuit Execution · qaoa.py QAOA Optimization
Components 18
Insurance Model
- Instance CSV Data
docs/data/YQH26_data - Bundling Model
code_examples/src/insurance_model.py
Classical Baseline
- ILP Solver
Challenge/Classical/classical.py - Baseline Results
Challenge/Classical/output/classical_baseline.json
DQI Approach
- Parity Checks
Challenge/DQI/dqi.py - DQI Circuit
Challenge/DQI/dqi.py - Circuit Execution
Challenge/DQI/dqi.py - Post-select and BP1
Challenge/DQI/dqi.py - Run Metrics
Challenge/DQI/dqi.py - DQI Results
Challenge/DQI/output
QAOA Approach
- Package QUBO
Challenge/QAOA/qaoa.py - QAOA Optimization
Challenge/QAOA/qaoa.py - Circuit Execution
Challenge/QAOA/qaoa.py - QAOA Results
Challenge/QAOA/output
Other components
- Experiment Runner
- PuLP / CBC
- Quantinuum Nexus
- Selene Emulator
Connections 20
- Experiment Runner → ILP Solver: runs baseline
- Experiment Runner → Parity Checks: runs DQI
- Experiment Runner → Package QUBO: runs QAOA
- Instance CSV Data → Bundling Model: loads instance
- ILP Solver → Bundling Model: loads and solves
- ILP Solver → PuLP / CBC: solves ILP
- ILP Solver → Baseline Results: writes baseline
- Bundling Model → Parity Checks: provides problem
- Parity Checks → DQI Circuit: builds checks
- DQI Circuit → Circuit Execution: provides circuits
- Circuit Execution → Quantinuum Nexus: submits circuits
- Circuit Execution → Post-select and BP1: returns measurements
- Post-select and BP1 → Run Metrics: evaluates solutions
- Run Metrics → DQI Results: writes results
- Bundling Model → Package QUBO: provides problem
- Package QUBO → QAOA Optimization: provides QUBO
- QAOA Optimization → Circuit Execution: evaluates parameters
- Circuit Execution → Selene Emulator: runs circuits
- QAOA Optimization → QAOA Results: writes results
- Baseline Results → QAOA Results: adds comparison