论文部分内容阅读
用科学的方法,根据有限的实验数据(样本观测值)正确估计和推断总体性质,这就是统计推断问题。统计推断从内容上可分为总体分布的参数估计和统计假设检验两类。例如,铸件的某些质量指标是随机的,要从实验观测值估计该指标的平均值(如强度、延伸率等)或指标的波动情况;有两种工艺方案,要正确判断其优劣来决定取舍。前者属于总体分布的参数估计,后者属于统计假设检验。统计推断是通过抽样(实验),由样本观测值(实验数据)建立的统计量去推断总体的相应
Using scientific methods to correctly estimate and infer overall properties based on limited experimental data (sample observations), this is the statistical inference problem. Statistical inference can be divided into two categories, the overall distribution of parameter estimation and statistical hypothesis testing. For example, some quality indicators of castings are random, and the average value (such as intensity, elongation, etc.) or index fluctuations of the index should be estimated from the experimental observations; there are two kinds of process schemes, and the merits must be correctly judged. Decide to choose. The former belongs to the parameter estimation of the overall distribution, and the latter belongs to the statistical hypothesis test. Statistical inference is the use of sampling (experimental) to infer the overall response from the statistics established by sample observations (experimental data).