🎓
An Evaluation System
  • An Evaluation System for Interpretable Machine Learning
  • 1. Abstract
  • 2. Introduction
  • 3. Interpretable Machine Learning
    • 3.1 Definition
    • 3.2 Methods
      • 3.2.1 Machine Learning Model
      • 3.2.2 Model-Agnostic
      • 3.2.3 Sample Theory
  • 4. Evaluation System of Interpretable Machine Learning
    • 4.1 Definition
    • 4.2 The Structure of the Evaluation System
      • 4.2.1 Interpretable Models(Model specific)
      • 4.2.2 Model Agnostic
      • 4.2.3 Human Explanations
    • 4.3 Reference Score Table
  • 5. Customized - Interpretable Machine Learning Model
    • 5.1 Why & How the user customize their model?
    • 5.2 Example
      • 5.2.1 Suggestion Lists
      • 5.2.2 Notebook
    • 5.3 Results & Explanation
  • 6. Discussion
  • 7. Citation and License
Powered by GitBook
On this page
  • 5.2.1.1 Model-Specific
  • 5.2.1.2 Model-Agnostic

Was this helpful?

  1. 5. Customized - Interpretable Machine Learning Model
  2. 5.2 Example

5.2.1 Suggestion Lists

5.2.1.1 Model-Specific

The following features are selected by the user:

  • Interaction

  • Parametric

  • Algorithmic complexity

According to the score table, we can get:

Interaction

Parametric

Algorithmic complexity

SUM

AVERAGE

SUGGESTION

GLM

1

1

1

3

2

STRONG

Gradient Boosting

0.7

1

0.8

2.5

2

STRONG

XGBoost

0.7

0.9

0.7

2.3

2

STRONG

Distributed Random Forest

0.7

1

0.9

2.6

2

STRONG

Deep Learning

0

0.1

0.2

0.3

2

LIGHTLY

K-means

0.1

0.7

0.5

1.3

2

MEDIUM

Therefore, the Model-Specific methods we suggest using are:

  • GLM

  • Gradient Boosting

  • XGBoost

  • Distributed Random Forest

5.2.1.2 Model-Agnostic

The following features are selected by the user:

  • Model internals

  • Datapoint

  • Detailed

  • Certainty

According to the score table, we can get:

Model internals

Datapoint

Detailed

Certainty

SUM

AVERAGE

SUGGESTION

Variable Importance / Standardized Coefficient

1

0.5

1

1

3.5

2.6

STRONG

PDP

0.1

0.9

0.5

0.8

2.3

2.6

MEDIUM

ICE

0.2

1

1

0.9

3.1

2.6

STRONG

ALE

0

0.9

0.6

0

1.5

2.6

LIGHTLY

Therefore, the Model-Agnostic methods we suggest using are

  • Variable Importance

  • PDP

  • ICE

Previous5.2 ExampleNext5.2.2 Notebook

Last updated 4 years ago

Was this helpful?