准确率，精准率和召回率（Accuracy、precision & Recall）

最重要分清四个变量：
https://en.wikipedia.org/wiki/Sensitivity_and_specificity

1. True positive (TP): e.g., Sick people correctly identified as sick
2. False positive (FP): e.g., Healthy people incorrectly identified as sick
3. True negative (TN): e.g., Healthy people correctly identified as healthy
4. False negative (FN): e.g., Sick people incorrectly identified as healthy

In general, Positive = identified and negative = rejected. Therefore:

1. True positive = correctly identified
2. False positive = incorrectly identified
3. True negative = correctly rejected
4. False negative = incorrectly rejected

所以：
condition positive （P)：the number of real positive cases in the data
condition negative (N)：the number of real negative cases in the data

即 P = TP + FN, F = TN + FP

---

准确率 Accuracy = (TP + TN) / (P + N) = (TP + TN) / (TP + FN + TN + FP)

精确率Precision = TP / (TP + FP)

召回率Recall = TP / P = TP / (TP + FN)

F1-Score = 2 * (Precision * Recall) / (Precision + Recall) = 2 * TP / (2 * TP + FP +FN)

---

示例说明：
1000个病人，现被确诊400个病人，600个健康人。
这400个病人实际上有300个病人（TP）和100给误判的健康人（FP）；
而这判定为健康的600人中，有500个是健康人（TN），100个病人（FN）。
即: TP = 300， FP = 100， TN = 500， FN = 100.

注：实际上的患者为500人（P = TP + FN），500个健康人（F = TN + FP）

所以:
准确率Accuray = (TP + TN) / (TP + FN + TN + FP) = 800 / 1000 = 80%
精确率Precision = TP / (TP + FP) = 300 / （300 + 100）= 75%
召回率Recall = TP / (TP + FN) = 300 / （300 + 100） = 75%
F1-score = 2 * (Precision * Recall) / (Precision + Recall) = 75%