|
In probability theory and statistics, a stochastic prediction procedure is based on a Bernoulli space and may be used to make predictions under specific conditions. In contrast to a prediction obtained in traditional science, predictions obtained by means of a stochastic predíction procedure〔Elart von Collani (ed.), ''Defining the Science of Stochastics'', Heldermann Verlag, Lemgo, 2004.〕 meet a given reliability requirement and are optimal with respect to accuracy. A prediction procedure refers to a random variable ''X'' and predicts future events for ''X'' depending on the initial conditions or more precisely said on what is known about the initial conditions. ==Mathematical formulation== A stochastic prediction procedure is a mathematical function〔Elart von Collani, The Need for a Standard for Making Predictions, ''Economic Quality Control'', Vol. 23, 287–299, 2008.〕 denoted defined on sets representing the possible initial conditions and having images that are the predictions. The function is derived meeting the following two requirements: * The stochastic prediction procedure shall yield predictions which will occur with a probability of at least where is called the reliability level of . * The stochastic prediction procedure shall yield predictions with optimal accuracy, where accuracy is defined by the size of the prediction. The mathematical task consists of deriving a function which meets the above formulated two requirements. 抄文引用元・出典: フリー百科事典『 ウィキペディア(Wikipedia)』 ■ウィキペディアで「Stochastic prediction procedure」の詳細全文を読む スポンサード リンク
|