论文部分内容阅读
昆虫种群动态的预测预报是昆虫种群数理生态学研究的重要方面之一。近些年来,不少数学分析方法和预报模型相继提出,列联表分析就是其中之一,如朱伯承提出的由列联表直接分析各因子与预报量的“单相关”,从而建立预报方程(下简称“综合相关法”)以及李火苟提出的用条件频率法简化《综合相关法》(下简称“条件频率法”)等。 本文将在上述工作的基础上,给出一种能充分、客观利用信息,试报满意,算法简便的以表征各因子相关程度的随机列联系数为权重,定义因子和预报量的条件频率之和的列联参数为预报依据的加权列联表因素分析方法,并举例说明之。
Prediction and prediction of insect population dynamics is one of the most important aspects of insect ecology research. In recent years, many mathematical analysis methods and forecasting models have been proposed one after another, and contingency table analysis is one of them. For example, Zhu Bocheng proposed a “single correlation” between each factor and the forecast quantity directly from the contingency table so as to establish a forecasting equation (Hereinafter referred to as “Comprehensive Correlation Law”) and Li Gou Gou put forward the use of conditional frequency method to simplify the “Comprehensive Correlation Law” (hereinafter referred to as “Conditional Frequency Method”). On the basis of the above work, this dissertation will give a condition frequency, which can fully and objectively use information, test report is satisfactory, and the algorithm is simple and convenient to represent the correlation coefficient of each factor as weight, definition factor and forecasting frequency And the associated parameters for the prediction based on the weighted contingency table analysis of factors, and an example.