Intrigued by the art of cloud architecture? Discover how to design, develop, and manage robust, secure, scalable, and dynamic solutions on Google Cloud as you prepare for the Professional Cloud Architect exam!
Prepare and test your skills
Prepare and test your skills
Worked example. The correct answer is already marked and every option is explained below, so there is nothing to select here. To answer questions yourself, start the free trial.
Keep the momentum going with these hand-picked practice scenarios
Want more questions like this?
Get a free certification question every week.
Last updated
An enterprise retail organization is preparing historical transaction data to train a tabular demand forecasting model in Vertex AI to predict daily sales per product across multiple store locations. The data engineering team is using Apache Beam on Dataflow to build a scalable data preprocessing pipeline that extracts, cleans, and structures raw records before importing them into Vertex AI.
Which data formatting and schema design must the Dataflow pipeline implement to produce a valid dataset for the forecasting model objective?
Format the dataset in a wide structure containing a categorical target column with distinct class labels, an entity timestamp column, and separate columns for each predicted future date horizon.
Format the dataset in a narrow (long) structure containing a numerical target column with no null values, a time column with consistent granularity, a time series identifier column, and column names that do not begin with an underscore.
Shard the tabular dataset into compressed TFRecord files containing nested JSON structs, and allow the target column to accept null values for unobserved historical intervals.
Pivot the dataset so that each unique product and store combination forms an individual feature column, replace all null values with numeric zero (0), and combine all records into a single continuous time series.
Format the dataset in a wide structure containing a categorical target column with distinct class labels, an entity timestamp column, and separate columns for each predicted future date horizon.
Format the dataset in a narrow (long) structure containing a numerical target column with no null values, a time column with consistent granularity, a time series identifier column, and column names that do not begin with an underscore.
Vertex AI tabular forecasting models require a specific schema and data structure to accurately track temporal dynamics across distinct entities over time. Unlike standard classification or regression models that typically favor wide tables, forecasting models require data organized in a narrow (long) format.
StudyCerts
StudyCerts_), and cannot begin with an underscore.Using Dataflow to clean, validate, and shape raw transactional records into this standardized narrow schema ensures full compatibility with Vertex AI's automated forecasting pipelines while preventing training errors.
Shard the tabular dataset into compressed TFRecord files containing nested JSON structs, and allow the target column to accept null values for unobserved historical intervals.
Pivot the dataset so that each unique product and store combination forms an individual feature column, replace all null values with numeric zero (0), and combine all records into a single continuous time series.