Inventory model with incomplete information: Sales and zero-balance signals

This study develops a periodic-review lost-sales inventory model in which sales are observed but inventory is revealed only when it reaches zero. A dynamic-programming approach simplifies belief updating and produces lower-bound approximations that improve when inventory variance is incorporated.

Alain Bensoussan, Metin Çakanyıldırım, Meng Li, Suresh Sethi (2025). Inventory model with incomplete information: Sales and zero-balance signals. Central European Journal of Operations Research 33, 571–584.