论文标题
当地信号的检测和估计
Detection and Estimation of Local Signals
论文作者
论文摘要
我们研究以一系列观测序列的级别，斜率或其他特性的变化点的形式检测和估算局部信号的最大分数统计量，并在似乎有多个变化时分段序列。我们发现，当观察结果串行依赖时，变化点会导致自相关的向上偏置估计，从而导致有时严重的功率损失。涉及温度变化的例子，大气温室气体的水平，自杀率，19009的发生率以及大流行期间的过量死亡说明了一般理论。
We study the maximum score statistic to detect and estimate local signals in the form of change-points in the level, slope, or other property of a sequence of observations, and to segment the sequence when there appear to be multiple changes. We find that when observations are serially dependent, the change-points can lead to upwardly biased estimates of autocorrelations, resulting in a sometimes serious loss of power. Examples involving temperature variations, the level of atmospheric greenhouse gases, suicide rates, incidence of COVID-19, and excess deaths during the pandemic illustrate the general theory.
