Classification Performance
TL;DR: The report analyzes the performance of a classification model
    Works for a single model or helps compare the two
    Works for binary and multi-class classification
    Displays a variety of plots related to the model performance
    Helps explore regions where the model makes different types of errors

Summary

Classification Performance report evaluates the quality of a classification model. It works both for binary and multi-class classification. If you have a probabilistic classification, refer to a separate report.
This report can be generated for a single model, or as a comparison. You can contrast your current production model performance against the past or an alternative model.

Requirements

To run this report, you need to have input features, and both target and prediction columns available.
You can use both numerical labels like "0", "1", "2" or class names like "virginica", "setoza", "versicolor" inside the target and prediction columns. The labels should be the same for the target and predictions.
NOTE: Column order in Binary Classification. For binary classification, class order matters. The tool expects that the target (so-called positive) class is the first in the column_mapping['prediction'] list.
To generate a comparative report, you will need the two datasets. The reference dataset serves as a benchmark. We analyze the change by comparing the current production data to the reference data.
You can also run this report for a single DataFrame , with no comparison performed. In this case, pass it as reference_data.

How it looks

The report includes 5 components. All plots are interactive.
At the moment, you cannot change or add custom quality metrics or plots. We expect to add this functionality in the future.

1. Model Quality Summary Metrics

We calculate a few standard model quality metrics: Accuracy, Precision, Recall, and F1-score.
To support the model performance analysis, we also generate interactive visualizations. They help analyze where the model makes mistakes and come up with improvement ideas.

2. Class Representation

Shows the number of objects of each class.

3. Confusion Matrix

Visualizes the classification errors and their type.

4. Quality Metrics by Class

Shows the model quality metrics for the individual classes.

5. Classification Quality by Feature

In this table, we show a number of plots for each feature. To expand the plots, click on the feature name.
In the tab “ALL”, we plot the distribution of classes against the values of the feature. This is the “Target Behavior by Feature” plot from the Categorial Target Drift report.
If you compare the two datasets, it visually shows the changes in the feature distribution and in the relationship between the values of the feature and the target.
Then, for each class, we plot the distribution of the True Positive, True Negative, False Positive, and False Negative predictions alongside the values of the feature.
It visualizes the regions where the model makes errors of each type and reveals the low-performance segments. This helps explore if a specific type of misclassification error is sensitive to the values of a given feature.

When to use the report

Here are our suggestions on when to use it—you can also combine it with the Data Drift and Categorical Target Drift reports to get a comprehensive picture. 1. To analyze the results of the model test. You can explore the results of an online or offline test and contrast it to the performance in training. Though this is not the primary use case, you can use this report to compare the model performance in an A/B test, or during a shadow model deployment. 2. To generate regular reports on the performance of a production model. You can run this report as a regular job (e.g. weekly or at every batch model run) to analyze its performance and share it with other stakeholders.
3. To analyze the model performance on the slices of data. By manipulating the input data frame, you can explore how the model performs on different data segments (e.g. users from a specific region).
4. To trigger or decide on the model retraining. You can use this report to check if your performance is below the threshold to initiate a model update and evaluate if retraining is likely to improve performance.
5. To debug or improve model performance. You can use the Classification Quality table to identify underperforming segments and decide on ways to address them.

JSON Profile

If you choose to generate a JSON profile, it will contain the following information:
1
{
2
"classification_performance": {
3
"name": "classification_performance",
4
"datetime": "datetime",
5
"data": {
6
"utility_columns": {
7
"date": null,
8
"id": null,
9
"target": "target",
10
"prediction": "prediction"
11
},
12
"cat_feature_names": [],
13
"num_feature_names": [],
14
"metrics": {
15
"reference": {
16
"accuracy": accuracy,
17
"precision": precision,
18
"recall": recall,
19
"f1": f1,
20
"metrics_matrix": {
21
"label_1": {
22
"precision": precision,
23
"recall": recall,
24
"f1-score": f1,
25
"support": support
26
},
27
"accuracy": accuracy,
28
"macro avg": {
29
"precision": precision,
30
"recall": recall,
31
"f1-score": f1,
32
"support": support
33
},
34
"weighted avg": {
35
"precision": precision,
36
"recall": recall,
37
"f1-score": f1,
38
"support": support
39
}
40
},
41
"confusion_matrix": {
42
"labels": [],
43
"values": []
44
]
45
}
46
},
47
"current": {
48
"accuracy": accuracy,
49
"precision": precision,
50
"recall": recall,
51
"f1": f1,
52
"metrics_matrix": {
53
"label_1": {
54
"precision": precision,
55
"recall": recall,
56
"f1-score": f1,
57
"support": support
58
},
59
"accuracy": accuracy,
60
"macro avg": {
61
"precision": precision,
62
"recall": recall,
63
"f1-score": f1,
64
"support": support
65
},
66
"weighted avg": {
67
"precision": precision,
68
"recall": recall,
69
"f1-score": f1,
70
"support": support
71
}
72
},
73
"confusion_matrix": {
74
"labels": [],
75
"values": []
76
]
77
}
78
},
79
}
80
}
81
},
82
"timestamp": "timestamp"
83
}
Copied!

Examples

    Browse our examples for sample Jupyter notebooks.
You can also read the release blog.
Last modified 3mo ago