Use AutoAI and model refinement capabilities to predict credit risk.ipynb

Kernel: Python 3

Use AutoAI and model refinement capabilities to predict credit risk with ibm-watsonx-ai

This notebook contains the steps and code to demonstrate support of AI Lifecycle features of am AutoAI model in watsonx.ai Runtime Service in watsonx.ai Runtime service. It contains steps and code to work with ibm-watsonx-ai library available in PyPI repository. It also introduces commands for training, persisting and deploying model, scoring it, updating the model and redeploying it.

Some familiarity with Python is helpful. This notebook uses Python 3.11.

Learning goals

The learning goals of this notebook are:

• Work with watsonx.ai Runtime experiments to train AutoAI models.

• Persist an external model in watsonx.ai Runtime repository.

• Deploy model for online scoring using client library.

• Score sample records using client library.

• Update previously persisted model.

• Redeploy model in-place.

Contents

This notebook contains the following parts:

1. Setup

2. Optimizer definition

3. Experiment Run

4. Persist and Deploy trained AutoAI model

5. Refine and persist new version of the model

6. Redeploy and score new version of the model

7. Clean up

8. Summary and next steps

1. Set up the environment

Before you use the sample code in this notebook, you must perform the following setup tasks:

• Create a watsonx.ai Runtime Service instance (a free plan is offered and information about how to create the instance can be found here).

• Create a Cloud Object Storage (COS) instance (a lite plan is offered and information about how to order storage can be found here).
  Note: When using Watson Studio, you already have a COS instance associated with the project you are running the notebook in.

Install and import the ibm-watsonx-ai and dependecies

Note: ibm-watsonx-ai documentation can be found here.

!pip install wget | tail -n 1
!pip install -U autoai-libs | tail -n 1
!pip install -U ibm-watsonx-ai | tail -n 1
!pip install -U 'snapml>=1.13.0,<1.14.0' | tail -n 1
!pip install -U 'lale[fairness]>=0.8.2,<0.9' | tail -n 1

Connection to watsonx.ai Runtime

Authenticate the watsonx.ai Runtime service on IBM Cloud. You need to provide Cloud API key and location.

Tip: Your Cloud API key can be generated by going to the Users section of the Cloud console. From that page, click your name, scroll down to the API Keys section, and click Create an IBM Cloud API key. Give your key a name and click Create, then copy the created key and paste it below. You can also get a service specific url by going to the Endpoint URLs section of the watsonx.ai Runtime docs. You can check your instance location in your watsonx.ai Runtime Service instance details.

You can use IBM Cloud CLI to retrieve the instance location.

ibmcloud login --apikey API_KEY -a https://cloud.ibm.com
ibmcloud resource service-instance INSTANCE_NAME

NOTE: You can also get a service specific apikey by going to the Service IDs section of the Cloud Console. From that page, click Create, and then copy the created key and paste it in the following cell.

Action: Enter your api_key and location in the following cell.

api_key = 'PUT_YOUR_KEY_HERE'
location = 'us-south'

from ibm_watsonx_ai import Credentials

credentials = Credentials(
    api_key=api_key,
    url='https://' + location + '.ml.cloud.ibm.com'
)

from ibm_watsonx_ai import APIClient

client = APIClient(credentials)

Working with spaces

You need to create a space that will be used for your work. If you do not have a space, you can use Deployment Spaces Dashboard to create one.

• Click New Deployment Space

• Create an empty space

• Select Cloud Object Storage

• Select watsonx.ai Runtime instance and press Create

• Copy space_id and paste it below

Tip: You can also use SDK to prepare the space for your work. More information can be found here.

Action: assign space ID below

space_id = 'PASTE YOUR SPACE ID HERE'

You can use the list method to print all existing spaces.

client.spaces.list(limit=10)

To be able to interact with all resources available in watsonx.ai Runtime, you need to set the space which you will be using.

client.set.default_space(space_id)

'SUCCESS'

Connections to COS

In next cell we read the COS credentials from the space.

cos_credentials = client.spaces.get_details(space_id=space_id)['entity']['storage']['properties']

2. Optimizer definition

Training data connection

Define connection information to COS bucket and training data CSV file. This example uses the German Credit Risk dataset.

The code in next cell uploads training data to the bucket.

filename = 'credit_risk_training_light.csv'
datasource_name = 'bluemixcloudobjectstorage'
bucketname = cos_credentials['bucket_name']

Download training data from git repository.

import os, wget

url = 'https://raw.githubusercontent.com/IBM/watson-machine-learning-samples/master/cloud/data/credit_risk/credit_risk_training_light.csv'
if not os.path.isfile(filename): wget.download(url)

Create connection

conn_meta_props= {
    client.connections.ConfigurationMetaNames.NAME: f"Connection to Database - {datasource_name} ",
    client.connections.ConfigurationMetaNames.DATASOURCE_TYPE: client.connections.get_datasource_type_id_by_name(datasource_name),
    client.connections.ConfigurationMetaNames.DESCRIPTION: "Connection to external Database",
    client.connections.ConfigurationMetaNames.PROPERTIES: {
        'bucket': bucketname,
        'access_key': cos_credentials['credentials']['editor']['access_key_id'],
        'secret_key': cos_credentials['credentials']['editor']['secret_access_key'],
        'iam_url': 'https://iam.cloud.ibm.com/identity/token',
        'url': cos_credentials['endpoint_url']
    }
}

conn_details = client.connections.create(meta_props=conn_meta_props)

Creating connections...
SUCCESS

Note: The above connection can be initialized alternatively with api_key and resource_instance_id. The above cell can be replaced with:

conn_meta_props= {
    client.connections.ConfigurationMetaNames.NAME: f"Connection to Database - {db_name} ",
    client.connections.ConfigurationMetaNames.DATASOURCE_TYPE: client.connections.get_datasource_type_id_by_name(db_name),
    client.connections.ConfigurationMetaNames.DESCRIPTION: "Connection to external Database",
    client.connections.ConfigurationMetaNames.PROPERTIES: {
        'bucket': bucket_name,
        'api_key': cos_credentials['apikey'],
        'resource_instance_id': cos_credentials['resource_instance_id'],
        'iam_url': 'https://iam.cloud.ibm.com/identity/token',
        'url': 'https://s3.us.cloud-object-storage.appdomain.cloud'
    }
}

conn_details = client.connections.create(meta_props=conn_meta_props)

connection_id = client.connections.get_id(conn_details)

Define connection information to training data.

from ibm_watsonx_ai.helpers import DataConnection, S3Location


credit_risk_conn = DataConnection(
    connection_asset_id=connection_id,
    location=S3Location(bucket=bucketname,
                        path=filename))

training_data_reference=[credit_risk_conn]

Check the connection information. Upload the data and validate.

credit_risk_conn._wml_client = client
credit_risk_conn.write(data=filename, remote_name=filename)
credit_risk_conn.read()

Optimizer configuration

Provide the input information for AutoAI optimizer:

• name - experiment name

• prediction_type - type of the problem

• prediction_column - target column name

• scoring - optimization metric

from ibm_watsonx_ai.experiment import AutoAI

experiment = AutoAI(credentials, space_id=space_id)

pipeline_optimizer = experiment.optimizer(
    name='Credit Risk Prediction - AutoAI',
    prediction_type=AutoAI.PredictionType.BINARY,
    prediction_column='Risk',
    scoring=AutoAI.Metrics.ROC_AUC_SCORE,
)

Configuration parameters can be retrieved via get_params().

pipeline_optimizer.get_params()

{'name': 'Credit Risk Prediction - AutoAI',
 'desc': '',
 'prediction_type': 'binary',
 'prediction_column': 'Risk',
 'prediction_columns': None,
 'timestamp_column_name': None,
 'scoring': 'roc_auc',
 'holdout_size': None,
 'max_num_daub_ensembles': None,
 't_shirt_size': 'l',
 'train_sample_rows_test_size': None,
 'include_only_estimators': None,
 'include_batched_ensemble_estimators': None,
 'backtest_num': None,
 'lookback_window': None,
 'forecast_window': None,
 'backtest_gap_length': None,
 'cognito_transform_names': None,
 'csv_separator': ',',
 'excel_sheet': None,
 'encoding': 'utf-8',
 'positive_label': None,
 'drop_duplicates': True,
 'outliers_columns': None,
 'text_processing': None,
 'word2vec_feature_number': None,
 'daub_give_priority_to_runtime': None,
 'text_columns_names': None,
 'sampling_type': None,
 'sample_size_limit': None,
 'sample_rows_limit': None,
 'sample_percentage_limit': None,
 'number_of_batch_rows': None,
 'n_parallel_data_connections': None,
 'test_data_csv_separator': ',',
 'test_data_excel_sheet': None,
 'test_data_encoding': 'utf-8',
 'categorical_imputation_strategy': None,
 'numerical_imputation_strategy': None,
 'numerical_imputation_value': None,
 'imputation_threshold': None,
 'retrain_on_holdout': True,
 'feature_columns': None,
 'pipeline_types': None,
 'supporting_features_at_forecast': None,
 'numerical_columns': None,
 'categorical_columns': None,
 'confidence_level': None,
 'incremental_learning': None,
 'early_stop_enabled': None,
 'early_stop_window_size': None,
 'time_ordered_data': None,
 'feature_selector_mode': None,
 'run_id': None}

3. Experiment run

Call the fit() method to trigger the AutoAI experiment. You can either use interactive mode (synchronous job) or background mode (asychronous job) by specifying background_model=True.

run_details = pipeline_optimizer.fit(
            training_data_reference=training_data_reference,
            background_mode=False)

Training job 81990bbc-6f9d-4101-bde6-985189c84071 completed: 100%|████████| [04:03<00:00,  2.43s/it]

You can use the get_run_status() method to monitor AutoAI jobs in background mode.

pipeline_optimizer.get_run_status()

'completed'

Load trained model

pipeline = pipeline_optimizer.get_pipeline()

Reading training data from COS

train_df = pipeline_optimizer.get_data_connections()[0].read()

train_X = train_df.drop(['Risk'], axis=1).values
train_y = train_df.Risk.values

4. Persist and Deploy trained AutoAI model

In this section you will learn how to deploy and score pipeline model as webservice using watsonx.ai Runtime instance.

Online deployment creation

from ibm_watsonx_ai.deployment import WebService

service = WebService(credentials, source_space_id=space_id)

service.create(
    experiment_run_id=run_details["metadata"]["id"],
    model="Pipeline_1", 
    deployment_name="AutoAI credit-risk deployment")

Preparing an AutoAI Deployment...
Published model uid: 6eb0393b-5175-4ea7-b60e-51696e64dfbf
Deploying model 6eb0393b-5175-4ea7-b60e-51696e64dfbf using V4 client.


#######################################################################################

Synchronous deployment creation for uid: '6eb0393b-5175-4ea7-b60e-51696e64dfbf' started

#######################################################################################


initializing
Note: online_url and serving_urls are deprecated and will be removed in a future release. Use inference instead.
..
ready


------------------------------------------------------------------------------------------------
Successfully finished deployment creation, deployment_uid='a7356206-c33a-4425-9ed9-a33fdd1f7d46'
------------------------------------------------------------------------------------------------

Deployment object could be printed to show basic information:

print(service)

To show all available information about the deployment use the .get_params() method:

service.get_params()

Scoring of webservice

You can make scoring request by calling score() on deployed pipeline.

predictions = service.score(payload=train_df.drop(['Risk'], axis=1).iloc[:10])
predictions

{'predictions': [{'fields': ['prediction', 'probability'],
   'values': [['No Risk', [1.0, 0.0]],
    ['No Risk', [1.0, 0.0]],
    ['No Risk', [1.0, 0.0]],
    ['No Risk', [1.0, 0.0]],
    ['Risk', [0.1, 0.9]],
    ['Risk', [0.0, 1.0]],
    ['No Risk', [1.0, 0.0]],
    ['No Risk', [1.0, 0.0]],
    ['No Risk', [0.9226527570789866, 0.07734724292101341]],
    ['Risk', [0.0, 1.0]]]}]}

If you want to work with the web service in an external Python application you can retrieve the service object by:

• Initialize the service by service = WebService(wml_credentials)

• Get deployment_id by service.list() method

• Get webservice object by service.get('deployment_id') method

After that you can call service.score() method.

Deleting deployment

You can delete the existing deployment by calling the service.delete() command. To list the existing web services you can use service.list().

5. Refine and persist new version of the model

In this section, you'll learn how to store new refinement version of your model in watsonx.ai Runtime repository by using the watsonx.ai Client.

Pipeline refinement with Lale

In this section you learn how to refine and retrain the best pipeline returned by AutoAI. There are many ways to refine a pipeline. For illustration, simply replace the final estimator in the pipeline by an interpretable model. The call to wrap_imported_operators() augments scikit-learn operators with schemas for hyperparameter tuning.

from sklearn.linear_model import LogisticRegression as LR
from sklearn.tree import DecisionTreeClassifier as Tree
from sklearn.neighbors import KNeighborsClassifier as KNN
from lale.lib.lale import Hyperopt
from lale import wrap_imported_operators

wrap_imported_operators()

Pipeline decomposition and new definition

Start by removing the last step of the pipeline, i.e., the final estimator.

prefix = pipeline.remove_last().freeze_trainable()
prefix.visualize()

Next, add a new final step, which consists of a choice of three estimators. In this code, | is the or combinator (algorithmic choice). It defines a search space for another optimizer run.

new_pipeline = prefix >> (LR | Tree | KNN)
new_pipeline.visualize()

New optimizer Hyperopt configuration and training

To automatically select the algorithm and tune its hyperparameters, we create an instance of the Hyperopt optimizer and fit it to the data.

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(train_X, train_y, test_size=0.15, random_state=33)

hyperopt = Hyperopt(estimator=new_pipeline, cv=3, max_evals=20, scoring='roc_auc')
hyperopt_pipelines = hyperopt.fit(X_train, y_train)

pipeline_model = hyperopt_pipelines.get_pipeline(astype='sklearn')

Save newly refined pipeline

import joblib

model_path = "refined_pipeline.pickle"
joblib.dump(pipeline_model, model_path)

import tarfile

with tarfile.open(model_path + ".tar.gz", "w:gz") as tar:
    tar.add(model_path)
model_path += ".tar.gz"

Publish new version of the model

Download current version of the model

model_id = service.get_params()['entity']['asset']['id']
deployment_id = service.get_params()['metadata']['id']

Note: online_url and serving_urls are deprecated and will be removed in a future release. Use inference instead.
Note: online_url and serving_urls are deprecated and will be removed in a future release. Use inference instead.

downloaded_model_name = "current_model.zip"
client.repository.download(model_id, downloaded_model_name)

import zipfile

with zipfile.ZipFile(downloaded_model_name, "r") as zip_ref:
    zip_ref.extractall()

Create archive containing updated model

updated_model_path = "updated_model.zip"
with zipfile.ZipFile(updated_model_path, "w") as zip_ref:
    zip_ref.write(model_path)
    zip_ref.write("pipeline-model.json")

Save the current model version

import json

print(json.dumps(client.repository.create_model_revision(model_id), indent=2))

Update the model

metadata = {
    client.repository.ModelMetaNames.NAME: 'AutoAI credit-risk updated model'
}

published_model = client.repository.update_model(
    model_id=model_id,
    update_model=updated_model_path,
    updated_meta_props=metadata
)

Save new model revision of the updated model

new_model_revision = client.repository.create_model_revision(model_id)
print(json.dumps(new_model_revision, indent=2))

Note: Model revisions can be identified by model id and rev number.

Get model rev number from creation details:

rev_id = new_model_revision['metadata'].get('rev')

List list existing revisions of the model

client.repository.list_models_revisions(model_id)

--  --------------------------------  ------------------------
ID  NAME                              CREATED
2   AutoAI credit-risk updated model  2024-03-07T10:02:23.516Z
1   P1                                2024-03-07T10:02:23.516Z
--  --------------------------------  ------------------------

Get details of updated model

print(json.dumps(client.repository.get_details(model_id), indent=2))

6. Redeploy and score new version of the model

In this section, you'll learn how to redeploy new version of the model by using the watsonx.ai Client.

metadata = {
    client.deployments.ConfigurationMetaNames.ASSET: {
        "id": model_id,
        "rev": rev_id
    }
}
    
updated_deployment = client.deployments.update(deployment_id, changes=metadata)

Note: online_url and serving_urls are deprecated and will be removed in a future release. Use inference instead.
Since ASSET is patched, deployment with new asset id/rev is being started. Monitor the status using deployments.get_details(deployment_uid) api

Wait for the deployment update:

import time

status = None
while status not in ['ready', 'failed']:
    print('.', end=' ')
    time.sleep(2)
    deployment_details = client.deployments.get_details(deployment_id)
    status = deployment_details['entity']['status'].get('state')
    
print("\nDeployment update finished with status: ", status)

. Note: online_url and serving_urls are deprecated and will be removed in a future release. Use inference instead.

Deployment update finished with status:  ready

Get updated deployment details

print(json.dumps(client.deployments.get_details(deployment_id), indent=2))

Score updated model

scoring_payload = {"input_data": [{"values": train_df.drop(['Risk'], axis=1).iloc[:5]}]}

Use client.deployments.score() method to run scoring.

predictions = client.deployments.score(deployment_id, scoring_payload)

print(json.dumps(predictions, indent=2))

{
  "predictions": [
    {
      "fields": [
        "prediction",
        "probability"
      ],
      "values": [
        [
          "No Risk",
          [
            1.0,
            0.0
          ]
        ],
        [
          "No Risk",
          [
            1.0,
            0.0
          ]
        ],
        [
          "No Risk",
          [
            1.0,
            0.0
          ]
        ],
        [
          "No Risk",
          [
            1.0,
            0.0
          ]
        ],
        [
          "Risk",
          [
            0.1,
            0.9
          ]
        ]
      ]
    }
  ]
}

7. Clean up

If you want to clean up all created assets:

• experiments

• trainings

• pipelines

• model definitions

• models

• functions

• deployments

please follow up this sample notebook.

8. Summary and next steps

You successfully completed this notebook! You learned how to use scikit-learn machine learning as well as watsonx.ai Runtime for model creation and deployment. Check out our Online Documentation for more samples, tutorials, documentation, how-tos, and blog posts.

Authors

Jan Sołtysik, Intern

Mateusz Szewczyk, Software Engineer at watsonx.ai.

Copyright © 2020-2025 IBM. This notebook and its source code are released under the terms of the MIT License.