A Bernoulli autoregressive moving average model applied to rainfall occurrence
A Bernoulli autoregressive moving average model applied to rainfall occurrence
复制标题
DOI：
10.1080/03610918.2018.1468453
复制
发表时间：
2019-10-21
影响因子：
0.9
通讯作者：
Ribeiro Diniz, Carlos Alberto
中科院分区：
文献类型：
--
作者：
Milani, Eder Angelo;Hartmann, Marcelo;Ribeiro Diniz, Carlos Alberto
在本文中，我们提出了一种新的二进制时间序列模型，该模型包含自回归滑动平均结构。所提出的模型是Garma模型的扩展，可用于计算感兴趣事件发生的预测概率，在这些概率依赖于近期内的先前观测的情况下。提出的模型被用来分析一个真实的数据集，该数据集只包含数据0和1，指示位于巴西圣保罗州中部地区的一个城市没有或存在降雨。
In this article, we propose a new model for binary time series involving an autoregressive moving average structure. The proposed model, which is an extension of the GARMA model, can be used for calculating the forecast probability of an occurrence of an event of interest in cases where these probabilities are dependent on previous observations in the near term. The proposed model is used to analyze a real dataset involving a series that contains only data 0 and 1, indicating the absence or presence of rain in a city located in the central region of Sao Paulo state, Brazil.