Databricks-Machine-Learning-Associate Databricks Certified Machine Learning Associate Exam Questions and Answers

To apply the Pandas UDFpredictto each record of a Spark DataFrame, you use themapInPandasmethod. This method allows the Pandas UDF to operate on partitions of the DataFrame as pandas DataFrames, applying the specified function (predictin this case) to each partition. The correct code completion to execute this is simplymapInPandas(predict), which specifies the UDF to use without additional arguments orincorrect function calls.References:

• PySpark DataFrame documentation (Using mapInPandas with UDFs).

The lack of improvement in model accuracy across evaluations suggests that the optimization algorithm might not be effectively exploring the hyperparameter space. Iterative optimization algorithms like Tree-structured Parzen Estimators (TPE) or Bayesian Optimization can adapt based on previous evaluations, guiding the search towards more promising regions of the hyperparameter space.

Changing the optimization algorithm can lead to better utilization of the information gathered during each evaluation, potentially improving the overall accuracy.

References:

• Hyperparameter Optimization with Hyperopt

To display visual histograms and summaries of the numeric features in a Spark DataFrame, the Databricks utility functiondbutils.data.summarizecan be used. This function provides a comprehensive summary, including visual histograms.

Correct code:

dbutils.data.summarize(spark_df)

Other options likespark_df.describe()andspark_df.summary()provide textual statistical summaries but do not include visual histograms.

References:

• Databricks Utilities Documentation

To identify which task is causing the failure in the job, the team should review the matrix view in the Job's runs. The matrix view provides a clear and detailed overview of each task's status, allowing the team to quickly identify which task failed. This approach ismore efficient than running each notebook interactively, as it provides immediate insights into the job's execution flow and any issues that occurred during the run.

References:

• Databricks documentation on Jobs: Jobs in Databricks

By creating a binary feature variable for each feature with missing values to indicate whether a value has been imputed, the data scientist can preserve information about the original state of the data. This approach maintains the integrity of the dataset by marking which values are original and which are synthetic (imputed). Here are the steps to implement this approach:

• Identify Missing Values:Determine which features contain missing values.
• Impute Missing Values:Continue with median imputation or choose another method (mean, mode, regression, etc.) to fill missing values.
• Create Indicator Variables:For each feature that had missing values, add a new binary feature. This feature should be '1' if the original value was missing and imputed, and '0' otherwise.
• Data Integration:Integrate these new binary features into the existing dataset. This maintains a record of where data imputation occurred, allowing models to potentially weight these observations differently.
• Model Adjustment:Adjust machine learning models to account for these new features, which might involve considering interactions between these binary indicators and other features.

References

• "Feature Engineering for Machine Learning" by Alice Zheng and Amanda Casari (O'Reilly Media, 2018), especially the sections on handling missing data.
• Scikit-learn documentation on imputing missing values: https://scikit-learn.org/stable/modules/impute.html

For large datasets with many variables, Spark ML distributes the training of a linear regression model using iterative optimization methods. Specifically, Spark ML employs algorithms such as Gradient Descent or L-BFGS (Limited-memory Broyden–Fletcher–Goldfarb–Shanno) to iteratively minimize the loss function. These iterative methods are suitable for distributed computing environments and can handle large-scale data efficiently by partitioning the data across nodes in a cluster and performing parallel updates.References:

• Spark MLlib Documentation (Linear Regression with Iterative Optimization).

The Databricks Runtime for Machine Learning includes pre-installed packages and libraries essential for machine learning and deep learning, including MLflow. To use it, the machine learning engineer can simply select an appropriate Databricks Runtime ML version from the "Databricks Runtime Version" dropdown menu while creating their cluster. This selection ensures that all necessary machine learning libraries, including MLflow, are pre-installed and ready for use, avoiding the need to manually install them each time.

References

• Databricks documentation on creating clusters: https://docs.databricks.com/clusters/create.html

The code block to compute the root mean-squared error (RMSE) for a linear regression model in Spark ML should use theRegressionEvaluatorclass withmetricNameset to "rmse". Given the schema ofpreds_dfwith columnspredictionandactual, the correct evaluator setup will specifypredictionCol="prediction"andlabelCol="actual". Thus, the appropriate code block (Option C in your list) that usesRegressionEvaluatorto compute the RMSE is the correct choice. This setup correctly measures the performance of the regression model using the predictions and actual outcomes from the DataFrame.References:

• Spark ML documentation (Using RegressionEvaluator to Compute RMSE).

The code block provided by the machine learning engineer will perform the desired inference when the Feature Store feature set was logged with the model at model_uri. This ensures that all necessary feature transformations and metadata are available for the model to make predictions. The Feature Store in Databricks allows for seamless integration of features and models, ensuring that the required features are correctly used during inference.

References:

• Databricks documentation on Feature Store: Feature Store in Databricks

To ensure reproducible training and test sets, writing the split data sets to persistent storage is a reliable approach. This allows you to consistently load the same training and test data for each model run, regardless of cluster reconfiguration or other changes in the environment.

Correct approach:

• Split the data.
• Write the split data to persistent storage (e.g., HDFS, S3).
• Load the data from storage for each model training session.

train_df, test_df = spark_df.randomSplit([0.8,0.2], seed=42) train_df.write.parquet("path/to/train_df.parquet") test_df.write.parquet("path/to/test_df.parquet")# Later, load the datatrain_df = spark.read.parquet("path/to/train_df.parquet") test_df = spark.read.parquet("path/to/test_df.parquet")

References:

• Spark DataFrameWriter Documentation

To determine the number of machine learning models that can be trained in parallel, we need to calculate the total number of combinations of hyperparameters. The given hyperparameter grid includes:

• Hyperparameter 1: [2, 5, 10] (3 values)
• Hyperparameter 2: [50, 100] (2 values)

The total number of combinations is the product of the number of values for each hyperparameter:3 (values of Hyperparameter 1)×2 (values of Hyperparameter 2)=63 (values of Hyperparameter 1)×2 (values of Hyperparameter 2)=6

With 3-fold cross-validation, each combination of hyperparameters will be evaluated 3 times. Thus, the total number of models trained will be:6 (combinations)×3 (folds)=186 (combinations)×3 (folds)=18

However, the number of models that can be trained in parallel is equal to the number of hyperparameter combinations, not the total number of models considering cross-validation. Therefore, 6 models can be trained in parallel.

References:

• Databricks documentation on hyperparameter tuning: Hyperparameter Tuning

One-hot encoding transforms categorical variables into a format that can be provided to machine learning algorithms to better predict the output. However, when done prematurely or universally within a feature repository, it can be problematic:

• Dimensionality Increase:One-hot encoding significantly increases the feature space, especially with high cardinality features, which can lead to high memory consumption and slower computation.
• Model Specificity:Some models handle categorical variables natively (like decision trees and boosting algorithms), and premature one-hot encoding can lead to inefficiency and loss of information (e.g., ordinal relationships).
• Sparse Matrix Issue:It often results in a sparse matrix where most values are zero, which can be inefficient in both storage and computation for some algorithms.
• Generalization vs. Specificity:Encoding should ideally be tailored to specific models and use cases rather than applied generally in a feature repository.

References

• "Feature Engineering and Selection: A Practical Approach for Predictive Models" by Max Kuhn and Kjell Johnson (CRC Press, 2019).

To retrieve the metadata description of a feature table created using the Feature Store Client (referred here asfs), the correct method involves callingget_tableon thefsclient with the table name as an argument, followed by accessing thedescriptionattribute of the returned object. The code snippetfs.get_table("new_table").descriptioncorrectly achieves this by fetching the table object for "new_table" and then accessing its description attribute, where the metadata is stored. The other options do not correctly focus on retrieving the metadata description.References:

• Databricks Feature Store documentation (Accessing Feature Table Metadata).

One reason that results can differ between sklearn and Spark ML decision trees, despite identical data and hyperparameters, is that Spark ML decision trees test binned feature values as representative split candidates. Spark ML uses a method called "quantile binning" to reduce the number of potential split points by grouping continuous features into bins. This binning process can lead to different splits compared to sklearn, which tests all possible split points directly. This difference in the splitting algorithm can cause variations in the resulting trees.References:

• Spark MLlib Documentation (Decision Trees and Quantile Binning).

To use the pandas API on Spark, the data scientist can run the following code block:

importpyspark.pandasasps df = ps.DataFrame(spark_df)

This code imports the pandas API on Spark and converts the Spark DataFramespark_dfinto a pandas-on-Spark DataFrame, allowing the data scientist to use familiar pandas functions for further feature engineering.

References:

• Databricks documentation on pandas API on Spark: pandas API on Spark

Tree of Parzen Estimators (TPE) is a sequential model-based optimization algorithm that selects hyperparameter values based on the outcomes of previous trials. It models the probability density of good and bad hyperparameter values and makes informed decisions about which hyperparameters to try next.

This approach contrasts with methods like random search and grid search, which do not use information from previous trials to guide the search process.

References:

• Hyperopt and TPE

To calculate the overall cross-validation root-mean-squared error (RMSE), you average the RMSE values obtained from each validation fold. Given the RMSE values of 10.0, 12.0, and 17.0 for the three folds, the overall cross-validation RMSE is calculated as the average of these three values:

Overall CV RMSE=10.0+12.0+17.03=39.03=13.0Overall CV RMSE=310.0+12.0+17.0​=339.0​=13.0

Thus, the correct answer is 13.0, which accurately represents the average RMSE across all folds.References:

• Cross-validation in Regression (Understanding Cross-Validation Metrics).

A train-validation split is often preferred over k-fold cross-validation (with k > 2) when computational efficiency is a concern. With a train-validation split, only two models (one on the training set and one on the validation set) are trained, whereas k-fold cross-validation requires training k models (one for each fold).

This reduction in the number of models trained can save significant computational resources and time, especially when dealing with large datasets or complex models.

References:

• Model Evaluation with Train-Test Split

In the context of Spark MLlib, an estimator refers to an algorithm which can be "fit" on a DataFrame to produce a model (referred to as a Transformer), which can then be used to transform one DataFrame into another, typically adding predictions or model scores. This is a fundamental concept in machine learning pipelines in Spark, where the workflow includes fitting estimators to data to produce transformers.

References

• Spark MLlib Documentation:https://spark.apache.org/docs/latest/ml-pipeline.html#estimators