Inventory Optimization | Python Simulation & Exact Validation by Markus MeighanInventory Optimization | Python Simulation & Exact Validation by Markus Meighan

Inventory Optimization | Python Simulation & Exact Validation

Portfolio project
Markus Meighan
Independent synthetic demonstration — not paid client work.

Decision

How much stock should a business order when demand is uncertain? I modeled a single-product selling period with sales revenue, purchase cost, salvage and unmet-demand penalties.

Mathematics and validation

inventory-simulation.pdf
Derived the exact critical-fractile optimum and checked it against 200,000 simulated Poisson demand scenarios. Common random numbers provide a paired comparison of policies. Tests cover the probability mass, marginal-profit identity, edge cases and seeded reproduction.

Result

Under the illustrative assumptions, ordering 104 rather than 100 units increases expected profit by $7.85 per period (0.65%), reduces stockout probability from 47.34% to 32.16%, and increases expected leftover stock. Demand sensitivity matters much more than the small base-case gain.

Deliverables and limits

Reproducible Python/NumPy code, policy results, sensitivity analysis and a three-page LaTeX report. This is a single-period model with known Poisson demand, not a real business forecast or a claim of commercial savings.

0 likes · 3 views · Nov 10, 2026

Inventory decisions under uncertainty: exact optimization, 200,000 simulations, paired comparisons and a LaTeX report.

Tags