Free Databricks Databricks-Machine-Learning-Professional Practice Exam with Questions & Answers | Set: 2

A model signature is a specification of the input and output schema of a model, which defines the names, data types, and shapes of the model inputs and outputs. Logging a model signature with an MLflow model has several benefits, such as:

• The signature can be used to document what type of data is expected as input and output by the model, which can help users and developers understand how to use the model correctly.
• The signature can be used to enforce schema validation when serving models via REST API, which can help prevent errors and ensure data quality. For example, Azure Machine Learning enforces compliance with the model signature for MLflow models that contain a signature1.
• The signature can be used to enable automatic conversion of input data to the appropriate format for the model, such as pandas DataFrame, numpy array, or PyTorch tensor. This can simplify the data preprocessing and inference steps and reduce the code complexity.

References:

• MLflow Models Documentation, Model Signature and Input Example, p. 6-7
• mlflow.models.signature Documentation, p. 1-2
• Guidelines for deploying MLflow models - Azure Machine Learning, p. 1

 The mlflow.register_model function takes two arguments: model_uri and name. The model_uri is the URI of the model that was logged to MLflow, which can be obtained from the best run object. The name is the name of the registered model in the MLflow Model Registry. Therefore, the correct line of code to register the model is:

mlflow.register_model(model_uri, “best_model”)

This will create a new registered model with the name “best_model” and register the model version from the best run as the first version of that model.

References:

• [mlflow.register_model — MLflow 1.22.0 documentation]
• [MLflow Model Registry — Databricks Documentation]
• [Manage MLflow Models — Databricks Documentation] Message has links.

The fs.read_table operation can be used to return a Spark DataFrame of a data set associated with a Feature Store table. This operation takes the name of the Feature Store table and an optional time travel specification as arguments. The fs.create_table operation is used to create a new Feature Store table from a Spark DataFrame. The fs.write_table operation is used to write data to an existing Feature Store table. The fs.get_table operation is used to get the metadata of a Feature Store table, not the data itself. There is a way to accomplish this task with fs, so option D is incorrect. References:

• Feature Store Client
• Feature Store Tables

Concept drift in machine learning and data mining refers to the change in the relationships between input and output data in the underlying problem over time. In other domains, this change maybe called “covariate shift,” “dataset shift,” or “nonstationarity.” Concept drift can affect the performance and accuracy of predictive models that assume a static relationship between input and output variables. Concept drift can be caused by various factors, such as changes in user behavior, environmental conditions, market trends, etc. Concept drift can be detected and handled by various methods, such as periodic retraining, online learning, ensemble methods, etc References:

• Concept drift - Wikipedia
• A Gentle Introduction to Concept Drift in Machine Learning
• Model Drift & Machine Learning: Concept Drift, Feature Drift, Etc.
• Data Drift vs. Concept Drift: What Is the Difference?

A two-way Chi-squared test is a statistical test that can be used to compare the frequencies of categorical variables across two or more groups. In this case, the engineer can use a two-way Chi-squared test to compare the frequencies of the categorical input variable values across different time periods (e.g., months, quarters, etc.) and see if there is a significant difference in the proportion of missing values. A significant difference would indicate that the missing values are not randomly distributed and may affect the performance of the machine learning model. References:

• Databricks Machine Learning Professional Exam Guide, Section 4: Solution and Data Monitoring, p. 5
• Databricks Machine Learning Documentation, Monitoring ML Models, Data Drift Detection, p. 3

Feature drift is a change in the distribution of the input data over time, which can affect the performance and accuracy of the model. Monitoring feature drift is important to ensure that the model is still valid and reliable for the current data. One simple and low-cost method of monitoring numeric feature drift is to track the summary statistics trends of the input features, such as mean, standard deviation, minimum, maximum, etc. These statistics can be computed easily and efficiently, and can provide a quick overview of the changes in the data distribution. If the summary statistics deviate significantly from the baseline values, it may indicate a feature drift. However, summary statistics trends may not capture all the nuances of the data distribution, such as outliers, multimodality, skewness, etc. Therefore, other methods, such as statistical tests or visualizations, may be needed to complement the summary statistics trends and provide a more comprehensive analysis of the feature drift123 References:

• Monitoring Feature Drift - Databricks
• Monitor feature attribution skew and drift | Vertex AI | Google Cloud
• Monitoring Model Drift - 6 Different Methodologies - Qualdo

An HTTP Webhook is a mechanism that allows you to register a callback that is triggered by an event, such as a model registry event. The callback is an HTTP request that is sent to a specified URL, which can invoke an action or a notification on another platform or service. An HTTP Webhook is required for use cases that involve integrating the model registry with external tools or workflows that are not supported by Databricks1.

Sending a message to a Slack channel when a model version transitions stages is a use case that requires the use of an HTTP Webhook. This is because Slack is an external platform that is not natively integrated with Databricks, and the model registry events are not directly accessible by Slack. Therefore, to send a message to a Slack channel, you need to register an HTTP Webhook that is triggered by the model registry event of interest, such as MODEL_VERSION_TRANSITIONED_STAGE. The HTTP Webhook then sends a request to the Slack API endpoint that corresponds to the channel and the message content2.

The other options are incorrect because:

• Option A: Starting a testing job when a new model is registered does not require the use of an HTTP Webhook, but rather a job registry webhook. A job registry webhook is a type of webhook that triggers a job in a Databricks workspace when a model registry event occurs. A job registry webhook can be created using the Databricks REST API or the Python client databricks-registry-webhooks on PyPI3.
• Option B: Updating data in a source table for a Databricks SQL dashboard when a model version transitions to the Production stage does not require the use of an HTTP Webhook, but rather a Databricks SQL trigger. A Databricks SQL trigger is a mechanism that allows you to execute a SQL query or a notebook when a specified condition is met, such as a time interval or a file arrival. A Databricks SQL trigger can be created using the Databricks SQL UI or the Databricks REST API4.
• Option C: Sending an email alert when an automated testing job fails does not require the use of an HTTP Webhook, but rather a job alert. A job alert is a feature that allows you to send an email notification when a job run meets a specifiedcondition, such as a failure, a timeout, or a success. A job alert can be created using the Jobs UI or the Databricks REST API5.
• Option D: None of these use cases require the use of an HTTP Webhook is incorrect, as option E does require the use of an HTTP Webhook. References: MLflow Model Registry Webhooks on Databricks, Streamline MLOps With MLflow Model Registry Webhooks, Databricks SQL Triggers, Job Alerts, [Slack API]

To view the data visualizations that are logged with each run, you can go to the Artifacts section of the MLflow Run page in Databricks. The Artifacts section shows the files and directories that are logged as artifacts for a run. You can browse the artifact hierarchy and preview the files, such as images, text, or HTML1. You can also download the artifacts or copy their URIs for further use2. The other options are incorrect because:

• Option A: The MLflow Model Registry Model page shows the information and metadata of a registered model, such as its name, description, versions, stages, and lineage. It does not show the data visualizations that are logged with each run3.
• Option B: The Artifacts section of the MLflow Experiment page shows the artifacts that are logged for an experiment, not for a specific run. It does not allow you to preview the files or browse the artifact hierarchy4.
• Option C: Logged data visualizations can be viewed in Databricks using the Artifacts section of the MLflow Run page1.
• Option E: There is no Figures section of the MLflow Run page in Databricks. The Figures section is only available in the open source MLflow UI, which shows the plots that are logged as figures for a run5. References: View run artifacts, Log, list, and download artifacts, Manage models, View experiment artifacts, Logging Visualizations with MLflow