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

4 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 Dimensionality Reduction and Feature Extraction 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