您好。老师

NO.PZ2021083101000007 问题如下 Seleperformanta from the cross- valition set confusion matrices is presentein Exhibit 1:Azarov anBector evaluate the taset XYZ performanmetrifor Confusion Matrices anC in Exhibit 1. Azarov says, “For Ganyme’s purposes, we shoulmost concernewith the cost of Type I errors. ”Baseon Exhibit 1, whiconfusion matrix monstrates the most favorable value of the performanmetric thbest aresses Azarov’s concern? A.Confusion Matrix B.Confusion Matrix C.Confusion Matrix A is correct. Precision is the ratio of correctly prectepositive classes to all prectepositive classes anis useful in situations where the cost of false positives or Type I errors is high. Confusion Matrix A hthe highest precision antherefore monstrates the most favorable value of the performanmetric thbest aresses Azarov’s concern about the cost of Type I errors.Confusion Matrix A ha precision score of 0.95, whiis higher ththe precision scores of Confusion Matrix B (0.93) anConfusion Matrix C (0.86).B is incorrebecause precision, not accuracy, is the performanmeasure thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix B monstrates the most favorable value for the accurascore (0.92), whiis higher ththe accurascores of Confusion Matrix A (0.91) anConfusion Matrix C (0.91). Accurais a performanmeasure thgives equweight to false positives anfalse negatives anis consireappropriate performanmeasure when the class stribution in the taset is equ(a balancetaset). However, Azarov is most concernewith the cost of false positives, or Type I errors, annot with finng the equilibrium between precision anrecall. Furthermore, taset XYZ hunequ(unbalance class stribution between positive sentiment annegative sentiment sentences.C is incorrebecause precision, not recall or F1 score, is the performanmeasure thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix C monstrates the most favorable value for the recall score (0.97), whiis higher ththe recall scores of Confusion Matrix A (0.87) anConfusion Matrix B (0.90). Recall is the ratio of correctly prectepositive classes to all actupositive classes anis useful in situations where the cost of false negatives, or Type II errors, is high. However, Azarov is most concernewith the cost of Type I errors, not Type II errors.F1 score is more appropriate (thaccuracy) when there is unequclass stribution in the taset anit is necessary to measure the equilibrium of precision anrecall. Confusion Matrix C monstrates the most favorable value for the F1 score (0.92), whiis higher ththe F1 scores of Confusion Matrix A (0.91) anConfusion Matrix B (0.91). Although taset XYZ hunequclass stribution between positive sentiment annegative sentiment sentences, Azarov is most concernewith the cost of false positives, or Type I errors, annot with finng the equilibrium between precision anrecall. 考点Mol Training - PerformanEvaluation 如果F1、Precision、recall、AUC值如果互相冲突怎么办，哪个优先级更高

Confusion Matrix B Confusion Matrix C A is correct. Precision is the ratio of correctly prectepositive classes to all prectepositive classes anis useful in situations where the cost of false positives or Type I errors is high. Confusion Matrix A hthe highest precision antherefore monstrates the most favorable value of the performanmetric thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix A ha precision score of 0.95, whiis higher ththe precision scores of Confusion Matrix B (0.93) anConfusion Matrix C (0.86). B is incorrebecause precision, not accuracy, is the performanmeasure thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix B monstrates the most favorable value for the accurascore (0.92), whiis higher ththe accurascores of Confusion Matrix A (0.91) anConfusion Matrix C (0.91). Accurais a performanmeasure thgives equweight to false positives anfalse negatives anis consireappropriate performanmeasure when the class stribution in the taset is equ(a balancetaset). However, Azarov is most concernewith the cost of false positives, or Type I errors, annot with finng the equilibrium between precision anrecall. Furthermore, taset XYZ hunequ(unbalance class stribution between positive sentiment annegative sentiment sentences. C is incorrebecause precision, not recall or F1 score, is the performanmeasure thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix C monstrates the most favorable value for the recall score (0.97), whiis higher ththe recall scores of Confusion Matrix A (0.87) anConfusion Matrix B (0.90). Recall is the ratio of correctly prectepositive classes to all actupositive classes anis useful in situations where the cost of false negatives, or Type II errors, is high. However, Azarov is most concernewith the cost of Type I errors, not Type II errors. F1 score is more appropriate (thaccuracy) when there is unequclass stribution in the taset anit is necessary to measure the equilibrium of precision anrecall. Confusion Matrix C monstrates the most favorable value for the F1 score (0.92), whiis higher ththe F1 scores of Confusion Matrix A (0.91) anConfusion Matrix B (0.91). Although taset XYZ hunequclass stribution between positive sentiment annegative sentiment sentences, Azarov is most concernewith the cost of false positives, or Type I errors, annot with finng the equilibrium between precision anrecall. 考点Mol Training - PerformanEvaluation 老师好，不太明白这题的意思，当valition的p低的时候，valition更有可能被认为是positive的吧……所以B的Type I Error概率就很大，最应该担心B。。。。难道不是么？

NO.PZ2021083101000007问题如下 Seleperformanta from the cross- valition set confusion matrices is presentein Exhibit 1:Azarov anBector evaluate the taset XYZ performanmetrifor Confusion Matrices anC in Exhibit 1. Azarov says, “For Ganyme’s purposes, we shoulmost concernewith the cost of Type I errors. ”Baseon Exhibit 1, whiconfusion matrix monstrates the most favorable value of the performanmetric thbest aresses Azarov’s concern? A.Confusion Matrix B.Confusion Matrix C.Confusion Matrix A is correct. Precision is the ratio of correctly prectepositive classes to all prectepositive classes anis useful in situations where the cost of false positives or Type I errors is high. Confusion Matrix A hthe highest precision antherefore monstrates the most favorable value of the performanmetric thbest aresses Azarov’s concern about the cost of Type I errors.Confusion Matrix A ha precision score of 0.95, whiis higher ththe precision scores of Confusion Matrix B (0.93) anConfusion Matrix C (0.86).B is incorrebecause precision, not accuracy, is the performanmeasure thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix B monstrates the most favorable value for the accurascore (0.92), whiis higher ththe accurascores of Confusion Matrix A (0.91) anConfusion Matrix C (0.91). Accurais a performanmeasure thgives equweight to false positives anfalse negatives anis consireappropriate performanmeasure when the class stribution in the taset is equ(a balancetaset). However, Azarov is most concernewith the cost of false positives, or Type I errors, annot with finng the equilibrium between precision anrecall. Furthermore, taset XYZ hunequ(unbalance class stribution between positive sentiment annegative sentiment sentences.C is incorrebecause precision, not recall or F1 score, is the performanmeasure thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix C monstrates the most favorable value for the recall score (0.97), whiis higher ththe recall scores of Confusion Matrix A (0.87) anConfusion Matrix B (0.90). Recall is the ratio of correctly prectepositive classes to all actupositive classes anis useful in situations where the cost of false negatives, or Type II errors, is high. However, Azarov is most concernewith the cost of Type I errors, not Type II errors.F1 score is more appropriate (thaccuracy) when there is unequclass stribution in the taset anit is necessary to measure the equilibrium of precision anrecall. Confusion Matrix C monstrates the most favorable value for the F1 score (0.92), whiis higher ththe F1 scores of Confusion Matrix A (0.91) anConfusion Matrix B (0.91). Although taset XYZ hunequclass stribution between positive sentiment annegative sentiment sentences, Azarov is most concernewith the cost of false positives, or Type I errors, annot with finng the equilibrium between precision anrecall. 考点Mol Training - PerformanEvaluation Azarov anBector evaluate the taset XYZ performanmetrifor Confusion Matrices anC in Exhibit 1. Azarov says, “For Ganyme’s purposes, we shoulmost concernewith the cost of Type I errors. ”Baseon Exhibit 1, whiconfusion matrix monstrates the most favorable value of the performanmetric thbest aresses Azarov’s concern?就是最关心的是type I error。然后题目问的是哪个confusion matrix能够最好的表现出Azarov的担忧，不过，显示出来的是最好的值？就是最让他可以不用担忧吗？感觉题目的表述好难理解。不用担忧就是Precision越接近于1越好，所以是A，是这样吗？

NO.PZ2021083101000007 Confusion Matrix B Confusion Matrix C A is correct. Precision is the ratio of correctly prectepositive classes to all prectepositive classes anis useful in situations where the cost of false positives or Type I errors is high. Confusion Matrix A hthe highest precision antherefore monstrates the most favorable value of the performanmetric thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix A ha precision score of 0.95, whiis higher ththe precision scores of Confusion Matrix B (0.93) anConfusion Matrix C (0.86). B is incorrebecause precision, not accuracy, is the performanmeasure thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix B monstrates the most favorable value for the accurascore (0.92), whiis higher ththe accurascores of Confusion Matrix A (0.91) anConfusion Matrix C (0.91). Accurais a performanmeasure thgives equweight to false positives anfalse negatives anis consireappropriate performanmeasure when the class stribution in the taset is equ(a balancetaset). However, Azarov is most concernewith the cost of false positives, or Type I errors, annot with finng the equilibrium between precision anrecall. Furthermore, taset XYZ hunequ(unbalance class stribution between positive sentiment annegative sentiment sentences. C is incorrebecause precision, not recall or F1 score, is the performanmeasure thbest aresses Azarov’s concern about the cost of Type I errors. Confusion Matrix C monstrates the most favorable value for the recall score (0.97), whiis higher ththe recall scores of Confusion Matrix A (0.87) anConfusion Matrix B (0.90). Recall is the ratio of correctly prectepositive classes to all actupositive classes anis useful in situations where the cost of false negatives, or Type II errors, is high. However, Azarov is most concernewith the cost of Type I errors, not Type II errors. F1 score is more appropriate (thaccuracy) when there is unequclass stribution in the taset anit is necessary to measure the equilibrium of precision anrecall. Confusion Matrix C monstrates the most favorable value for the F1 score (0.92), whiis higher ththe F1 scores of Confusion Matrix A (0.91) anConfusion Matrix B (0.91). Although taset XYZ hunequclass stribution between positive sentiment annegative sentiment sentences, Azarov is most concernewith the cost of false positives, or Type I errors, annot with finng the equilibrium between precision anrecall. 考点Mol Training - PerformanEvaluation p值高和拒真怎么联系？