Difference between revisions of "Orange: Calibrated Learner"

From OnnoWiki
Jump to navigation Jump to search
 
(6 intermediate revisions by the same user not shown)
Line 1: Line 1:
 
Sumber: https://docs.biolab.si//3/visual-programming/widgets/model/calibratedlearner.html
 
Sumber: https://docs.biolab.si//3/visual-programming/widgets/model/calibratedlearner.html
  
−
Wraps another learner with probability calibration and decision threshold optimization.
+
Membungkus / melanjutkan kerja dari learner lain dengan probability calibration dan decision threshold optimization.
  
−
Inputs
+
==Input==
  
−
    Data: input dataset
+
Data: input dataset
−
    Preprocessor: preprocessing method(s)
+
Preprocessor: preprocessing method(s)
−
    Base Learner: learner to calibrate
+
Base Learner: learner to calibrate
  
−
Outputs
+
==Output==
  
−
    Learner: calibrated learning algorithm
+
Learner: calibrated learning algorithm
−
    Model: trained model using the calibrated learner
+
Model: trained model using the calibrated learner
  
−
This learner produces a model that calibrates the distribution of class probabilities and optimizes decision threshold. The widget works only for binary classification tasks.
+
Widget Calibrated Learner menghasilkan sebuah model yang mengkalibrasi distribusi dari class probabilities dan meng-optimasi decision threshold. Widget ini hanya bekerja untuk binary classification task saja.
  
 
[[File:Calibrated-Learner-stamped.png|center|200px|thumb]]
 
[[File:Calibrated-Learner-stamped.png|center|200px|thumb]]
  
  
−
    The name under which it will appear in other widgets. Default name is composed of the learner, calibration and optimization parameters.
+
* The name under which it will appear in other widgets. Default name is composed of the learner, calibration and optimization parameters.
 +
* Probability calibration:
 +
** Sigmoid calibration
 +
** Isotonic calibration
 +
** No calibration
  
−
    Probability calibration:
+
* Decision threshold optimization:
−
        Sigmoid calibration
+
** Optimize classification accuracy
−
        Isotonic calibration
+
** Optimize F1 score
−
        No calibration
+
** No threshold optimization
  
−
    Decision threshold optimization:
+
* Press Apply to commit changes. If Apply Automatically is ticked, changes are committed automatically.
−
        Optimize classification accuracy
 
−
        Optimize F1 score
 
−
        No threshold optimization
 
  
−
    Press Apply to commit changes. If Apply Automatically is ticked, changes are committed automatically.
+
==Contoh==
  
−
==Contoh==
+
Contoh sederhana dengan Calibrated Learner. Kita menggunakan dataset titanic karena widget ini membutuhkan nilai binary class (dalam hal ini mereka adalah 'survived’ atau ‘not survived’).
  
−
A simple example with Calibrated Learner. We are using the titanic data set as the widget requires binary class values (in this case they are ‘survived’ and ‘not survived’).
+
Kita menggunakan Logistic Regression sebagai base learner yang akan dikalibrasi dengan nilai setting default, yaitu dengan sigmoid optimization dari distribusi nilai dan di optimasi dengan CA.
  
−
We will use Logistic Regression as the base learner which will we calibrate with the default settings, that is with sigmoid optimization of distribution values and by optimizing the CA.
+
Membandingkan hasil dari uncalibrated Logistic Regression model kita akan melihat dengan jelas bahwa calibrated model lebih baik.
  
−
Comparing the results with the uncalibrated Logistic Regression model we see that the calibrated model performs better.
+
[[File:Calibrated-Learner-Example.png|center|600px|thumb]]
  
−
[[File:Calibrated-Learner-Example.png|center|200px|thumb]]
+
==Youtube==
  
 +
* [https://www.youtube.com/watch?v=fYLD2AHjjZI YOUTUBE: ORANGE model Callibrated Learner]
  
 
==Referensi==
 
==Referensi==

Latest revision as of 05:55, 11 April 2020

Sumber: https://docs.biolab.si//3/visual-programming/widgets/model/calibratedlearner.html

Membungkus / melanjutkan kerja dari learner lain dengan probability calibration dan decision threshold optimization.

Input

Data: input dataset
Preprocessor: preprocessing method(s)
Base Learner: learner to calibrate

Output

Learner: calibrated learning algorithm
Model: trained model using the calibrated learner

Widget Calibrated Learner menghasilkan sebuah model yang mengkalibrasi distribusi dari class probabilities dan meng-optimasi decision threshold. Widget ini hanya bekerja untuk binary classification task saja.

Calibrated-Learner-stamped.png


  • The name under which it will appear in other widgets. Default name is composed of the learner, calibration and optimization parameters.
  • Probability calibration:
    • Sigmoid calibration
    • Isotonic calibration
    • No calibration
  • Decision threshold optimization:
    • Optimize classification accuracy
    • Optimize F1 score
    • No threshold optimization
  • Press Apply to commit changes. If Apply Automatically is ticked, changes are committed automatically.

Contoh

Contoh sederhana dengan Calibrated Learner. Kita menggunakan dataset titanic karena widget ini membutuhkan nilai binary class (dalam hal ini mereka adalah 'survived’ atau ‘not survived’).

Kita menggunakan Logistic Regression sebagai base learner yang akan dikalibrasi dengan nilai setting default, yaitu dengan sigmoid optimization dari distribusi nilai dan di optimasi dengan CA.

Membandingkan hasil dari uncalibrated Logistic Regression model kita akan melihat dengan jelas bahwa calibrated model lebih baik.

Calibrated-Learner-Example.png

Youtube

Referensi

Pranala Menarik

Retrieved from "https://onnocenter.or.id/wiki/index.php?title=Orange:_Calibrated_Learner&oldid=61090"