Model Drift Detection

When ML models are deployed in production, sometimes even minor changes in a data distribution can adversely affect the performance of ML models. When the input data distribution shifts then prediction quality can drop. It is important to track this drift. This demo is based on the mixed-type tabular data drift detection method in the alibi detect project for tabular datasets.

Here we will :

  • Launch an income classifier model based on demographic features from a 1996 US census. The data instances contain a person’s characteristics like age, marital status or education while the label represents whether the person makes more or less than $50k per year.

  • Setup a mixed-type tabular data drift detector for this particular model.

  • Make a batch of predictions over time

  • Track the drift metrics in the Monitoring dashboard.

Important

This demo requires Knative installation on the cluster as the drift detector will be installed as a kservice. See Knative intallation instructions for necessary setup required.

Register an income classifier model

Register a pre-trained income classifier SKLearn model with model artefacts stored publicly in this Google bucket: gs://seldon-models/sklearn/income/model-0.23.2

  1. In the Model Catalogue page, click the Register New Model button:

    Expand to see the 'Register New Model' button

    register new model button

  2. In the Register New Model wizard, enter the following information, then click REGISTER MODEL:

    • Model Name: income-classifier

    • URI: gs://seldon-models/sklearn/income/model-0.23.2

    • Artifact Type: SciKit Learn

    • Version: v1

    Expand to see model configuration../../../../_images/configure-model.png

Configure predictions schema

Edit the model metadata to update the prediction schema for the model. The prediction schema is a generic schema structure for machine learning model predictions. It is a definition of feature inputs and output targets from the model prediction. Use the income classifier model predictions schema to edit and save the model level metadata. Learn more about the predictions schema at the ML Predictions Schema open source repository.

  1. Click on the model income-classifier model that you have just registered.

    Expand to see select model

    select model

  2. Click the Edit Metadata button to update the prediction schema associated with the model

  3. Paste the prediction schema and click Save Metadata.

    Expand to see configure prediction schema../../../../_images/configure-prediction-schema.png

Launch a Seldon Core deployment

Deploy the income classifier model from the catalogue into an appropriate namespace

  1. From the model catalogue, under the Action dropdown list, select Deploy.

    Expand to see deploy model

    deploy model

  2. Enter the deployment details in the deployment creation wizard and click Next:

    • Name: income-classifier

    • Type: Seldon Deployment

    • Protocol: Seldon

    Expand to see deploy model

    In the deployment creation wizard, enter a name for your new deployment (e.g. income-classifier). Select the namespace you would like the deployment to reside in (e.g. seldon). From the protocol dropdown menu, select Seldon and click Next. deployment creation wizard 1

  3. The predictor details should already be filled in from the model catalogue. Click Next:

    Expand to see default predictor details

    deployment creation wizard 2

  4. Click Next for the remaining steps, then click Launch.

Add A Drift Detector

From the deployment overview page, select your deployment to enter the deployment dashboard. Inside the deployment dashboard, add a drift detector with by clicking the Create button within the Drift Detection widget.

Expand to see drift detector creation

configure drift detector

create drift detector

Enter the following parameters in the modal popup which appears, to configure the detector:

  • Model Name: income-drift.

  • Model URI: (For public google buckets, secret field is optional)

    gs://seldon-models/alibi-detect/cd/tabular/income-0_8_1/
    
  • Reply URL: (By default, the Reply URL is set as seldon-request-logger in the logger’s default namespace. If you are using a custom installation, please change this parameter according to your installation.)

    http://seldon-request-logger.seldon-logs
    
  • Batch Size: 200

Then, click Create Drift-Detector to complete the setup.

The drift detector model used here was trained similarly to the mixed type data drift detection model in the alibi-detect documentation, with one key difference - the reference dataset used was a filtered for working-class adults where only entries with Age > 30(more than 30) and Hours per Week > 40(more than 40) were kept from the data set as reference.

Run Batch Predictions

  1. From the deployment dashboard, click on Batch Jobs. Run a batch prediction job using the ndarray payload format text predictions data file. This file has 4000 individual data points and based on our drift detector configuration, drift will be detected for a batch every 200 points. The distribution of the data in the first half section is the same as the distribution of the reference data the drift detector was configured with and the second half section of the data should be different to observe drift.

  2. Upload the data to a bucket store of your choice. This demo will use minio and store the data at bucket path minio://income-batch-data/data.txt . Do not forget to configure your storage access credentials secret - we have it as minio-bucket here. Refer to the batch request demo for an example of how this can be done via the minio browser.

  3. Running a batch job with the configuration below. This runs an offline job that makes a prediction request for a batch of 200 rows in the file at minio://income-batch-data/data.txt every 5 seconds:

    Input Data Location: minio://income-batch-data/data.txt
    Output Data Location: minio://income-batch-data/output.txt
    Number of Workers: 1
    Number of Retries: 3
    Batch Size: 200
    Minimum Batch Wait Interval (sec): 5
    Method: Predict
    Transport Protocol: REST
    Input Data Type: ndarray
    Object Store Secret Name: minio-bucket
    
Batch Job Details

Monitor Drift Detection Metrics

Under the Monitor section of your deployment navigation, on the Drift Detection Tab, you can see a timeline of drift detection metrics.

As mentioned above, the drift detection model was trained with data for working-class adults where only entries with Age > 30(more than 30) and Hours per Week > 40(more than 40) were kept from the data set as reference. The live prediction data set provided for the batch prediction jobs contains entries with Age > 30 and Hours per Week > 40 in the first 2000 rows referring to working class adults, and later entries with Age < 20(less than 20) and Hours per Week < 20(less than 20) in the last 2000 rows referring to teen aged part-time workers/students.

The drift dashboard showcases 2 types of metrics graphs:

  1. P-value score over time

    1. Zoomed in view, focussing on features that have drifted, i.e. features that have a p-value score of less than the threshold. p-values drift metrics zoomed in

    2. Zoomed out view, showing all features p-values drift metrics zoomed out

  2. Distance score over time. distance score metrics

Note that for both drift metrics graphs, the starting batches do not drift and are marked by an empty O symbol, while the later batches do drift and are marked by a filled O symbol.

Data drift and reference distributions comparison

To further analyse prediction data drift, you can also switch to the feature distribution tab to compare predictions to reference data distribution. See feature distribution monitoring demo for setup details.

Upload the income classifier reference dataset as the reference data to monitor data drift in terms of feature distributions. Once reference data is available, you can compare the distributions of the prediction data to the reference data.

Expand to see reference data available

You can see when reference data is available by checking the button on the top left of the DISTRIBUTIONS dashboard. If it is not clickable and displays REFERENCE DATA AVAILABLE, then reference data is available.

Reference data available button

When switching to the DISTRIBUTIONS dashboard, you may also see the following notification pop up:

Reference data available notification

monitor distributions

Troubleshooting

If you experience issues with this demo, see the troubleshooting docs and also the Knative or Elasticsearch sections.