EXPLORATION VERSUS EXPLOITATION TRADEOFF IN THE PARTIAL FEEDBACK PARADIGM IN DECISION MAKING FROM EXPERIENCE.

When facing two options with unknown characteristics, we have to explore to learn the properties. This gives two options: settle down on one choice, or explore both to find the true probabilities. This is known as the exploration-exploitation trade-off. In this research it is found that small-sample problems result in the distortion of probabilities, and settling down behavior is affected by experienced variance and the cost of learning. The hot stove effect and the gamblers fallacy are present. No conclusions on choice behavior in different type of problems is found, leaving room for the analysis of risk preferences in the trade-off.