دانلود کتاب تحلیل تصمیم: مدل‌های ریاضی و الگوریتم‌هایی برای تصمیم‌گیری‌های متوالی

Sequential decision-making is a fundamental part of the human experience. The ability to think and act rationally over time to achieve an objective can be broken down into the following steps: observe the available data, process it logically, make a decision based on the processing, and repeat this process. This can be summarized succinctly by the question: What shall I do next? Sequential decision-making arises in many contexts, including personal decisions, business operations, energy systems, finance, healthcare delivery, transportation systems, sports, and many other situations where a decision maker needs to select among multiple choices to decide what to do next. In some situations, the ”What next?” question is easy and determined by others’ preferences or expectations (e.g., customer expectations, employer policies). In other cases, the outcome of alternative decisions may be inconsequential, in which case the decision maker may be satisfied with several decisions and happy to let chance dictate what happens next (e.g., _lip a coin). However, decision-makers quickly become overwhelmed when the stakes are high and complicating factors are present. For instance, decision-making is challenging when allocating scarce or expensive resources, when there are numerous alternatives, and when there is uncertainty about the future. Complexity is high, especially when present decisions have implications for future decisions. People have been making decisions since the dawn of humanity, collecting data through personal experience or learning from others (e.g., parents, peers, teachers), and using their acquired knowledge to inform their decisions. However, it is only in the last century that computers and algorithms have significantly aided decision-makers. The first half of the 20th century witnessed significant advances across key areas essential to informed decision-making, including computer science, economics, mathematics, psychology, and statistics. The origins of operations research, a multidisciplinary field that blends these key areas, are often associated with military decision-making during World War II, when scientists were recruited to apply approaches from multiple fields to analyze data and recommend decisions on resource allocation, logistics planning, risk mitigation, and more. Gass and Assad provide a brief history of the field for interested readers. Since its early days, operations research and the associated scientific and mathematical approaches to decision-making have continually evolved to address increasingly complex decision-making situations. The progression of modern computing technology began in the 1950s with the creation of mainframes, followed by a steady stream of advances, including personal computers, laptops, and increasingly compact computing devices, such as tablets and smartphones. Over time, advances in computational speed, data storage efficiency, and cost have enabled continuous improvements in algorithms that automatically store and process enormous datasets, employing mathematical decision-making models that far surpass human capabilities in many tasks. However, this raw computing power is only as useful as the models and algorithms that transform data into decisions; thus, human decision-makers‘ judgment remains as important as ever in guiding the appropriate use of computing technology.