Modeling and forecasting are impressive and active research areas, which have been widely used in diverse theoretical and practical applications, successfully. Accuracy is the most important and at the same time is the most challenging feature of modeling and forecasting approaches that considerable efforts have been made in the literature for improving it. However, despite all interminable endeavors done in this regard, improving accuracy is yet often a problematic task. Weighting are among the most fundamental and critical parts of modeling and forecasting approaches, which have a significant impact on the performance and accuracy. Nevertheless, despite of the substantial importance of weighting algorithms, they have not been appropriately/comprehensively investigated in the literature. This fact that there are no such comprehensive review papers in this field is the core literature gap of modelling and forecasting that construct the main contribution of this paper. In this way, the main core objective of this paper is to comprehensively review and classify the most frequently applied weighting algorithms in various modeling and forecasting approaches. Motivated by this goal, in this review paper, different weighting algorithms, commonly used in the modelling and forecasting field, are classified based on the type of their function (i.e., Continuous/Discrete and Static/Dynamic) into four main categories. After that, the most related works done in these categorizes, including (1) discrete and static, (2) discrete and dynamic, (3) continuous and static, and (4) continuous and dynamic, are systematically reviewed to offer practical as well as theoretical guides for researchers. For this purpose, in each category, the most popular weighting algorithms, their potential limitations, and some highlighted advantages/disadvantages of them are briefly discussed. The weighting reviewed works, investigated in this paper, totally contain 350 paper that have been published since 2016 up to 2020. This is the first study that provides a comprehensive overview of the most frequently used weighting approaches employed in modeling and forecasting. The deep analysis and classification represented in this review paper revealed that the modeling and forecasting methods come from a unique unit which is named the weighting approach.
数据挖掘和知识发现中的加权方法：回顾,Neural Processing Letters
