混淆矩阵计算器
使用混淆矩阵导出的指标评估分类模型的性能，包括准确率、精确率、召回率和特异性。
混淆矩阵计算器 gives you a faster way to work through practical calculation scenarios without rebuilding the same calculation from scratch every time. Start with 真正例 (TP), 假正例 (FP), 真负例 (TN), and 假负例 (FN) and use the live outputs to estimate 特异性, 召回率 (敏感度), and 精确率 in one pass. This page is built to help you compare scenarios quickly, validate rough assumptions, and move into a related calculator when you need a second angle on the same problem.
Use The Calculator In A Few Steps
Work through the inputs deliberately, then compare outputs instead of relying on a single scenario.
Scenario Checks
Use these examples to verify the behavior of the calculator before you run your own values.
What Users Usually Need Next
Keep the calculator open while you skim these answers so you can test the scenario that matters.
Related Calculators
These links are chosen to deepen the session and answer the next likely question without sending the user back to search.
Mann-Whitney U 检验计算器 (Wilcoxon 秩和检验)
当正态性假设被违反时，进行非参数检验以比较两个独立组之间的差异。它计算 U 统计量和相关的 p 值。
P 值计算器
计算观察到的检验统计量与计算出的统计量一样极端或更极端的概率，假设原假设为真。
正态分布计算器
计算与标准正态 (Z) 分布或任何正态分布相关的概率和百分位数。
T检验计算器
执行t检验（单样本、双样本或配对）以比较均值并确定统计学意义。