how to manually adjust threshold value in PERFCURVE(​LABELS,SCO​RES,POSCLA​SS)

1 次查看（过去 30 天）
Balaji M. Sontakke
评论： Mohammad Sami 2020-3-5
I have 100 subjects, I used KNN classifier for classification, when i use perfcurve function for roc it automatically displays result, the question in my mind is how and where to adjust threshold values in perfcurve that causes on result.
  4 个评论
显示 2更早的评论
Balaji M. Sontakke
Balaji M. Sontakke 2020-3-3
Sir, as you said "You can then choose the optimal value of T from this array based on your needs.", But in my perfcurve, I get 11 values start from 1,2,3......10.
many questions are comes in mind regarding this threshold, I will ask one by one here...
1) How the researchers get threshold like in following table but in my result only I get threshold like 1,2,3....10
Threshold FAR (%) FRR (%) EER (%) GAR (%)
0.1 15.61 42.06 28.84 71.15
0.13 20.30 22.66 21.48 78.51
0.2 31.23 8.73 19.98 80.01
0.3 46.85 2.93 24.89 75.10
0.4 62.46 2 32.23 67.76
0.5 78.08 1.4 39.74 60.25
Following is the result of my program...
Confusion matrix
4 0 0 0 0 0 0 0 0 0
0 3 0 0 0 0 0 1 0 0
1 1 0 0 0 1 0 0 1 0
0 1 1 2 0 0 0 0 0 0
0 0 0 1 1 0 0 0 1 1
1 0 0 0 0 0 1 2 0 0
2 0 0 0 0 0 2 0 0 0
1 0 0 0 0 1 0 2 0 0
1 0 0 1 0 0 0 0 2 0
0 1 2 0 0 0 0 0 0 1
Total Instance = 40
class1==>1
class2==>2
class3==>3
class4==>4
class5==>5
class6==>6
class7==>7
class8==>8
class9==>9
class10==>10
Multi-Class Confusion Matrix Output
TruePositive FalsePositive FalseNegative TrueNegative
____________ _____________ _____________ ____________
Actual_class1 4 6 0 30
Actual_class2 3 3 1 33
Actual_class3 0 3 4 33
Actual_class4 2 2 2 34
Actual_class5 1 0 3 36
Actual_class6 0 2 4 34
Actual_class7 2 1 2 35
Actual_class8 2 3 2 33
Actual_class9 2 2 2 34
Actual_class10 1 1 3 35
Overall values
Accuracy: 0.4250
Error: 0.5750
Sensitivity: 0.4250
Specificity: 0.9361
Precision: 0.4467
FalsePositiveRate: 0.0639
F1_score: NaN
MatthewsCorrelationCoefficient: 0.3900
Kappa: 0.6870
Tbl =
11×3 table
FPR TPR Thr
_______ ___ ___
0 0 10
0.13333 0 10
0.23333 0.1 9
0.33333 0.2 8
0.4 0.4 7
0.5 0.5 6
0.63333 0.5 5
0.76667 0.5 4
0.86667 0.6 3
1 0.6 2
1 1 1

请先登录，再进行评论。

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Classification 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by