Optimization Under Incomplete Information: A Systematic Review of Automated Decision-Making in the Sequential Battleship Game

Decision-making under incomplete information is a recurrent challenge in optimization and artificial intelligence, requiring methods capable of reasoning under uncertainty and partial observability. In this context, the Sequential Battleship problem emerges as an interesting benchmark environment due to its combination of hidden states, probabilistic reasoning, sequential observations, and uncertainty reduction through information gathering. However, existing works are dispersed across different research areas, making systematic analysis and comparison difficult. This paper presents a Systematic Literature Review based on the PRISMA methodology to investigate automated decision-making approaches applied to the Sequential Battleship problem. The review analyzes the methodological evolution of the literature and the main computational paradigms adopted to address sequential optimization under uncertainty. A total of 37 studies were identified and categorized by methodological paradigm. Finally, the review discusses current limitations and highlights directions for future research.