Answer-first summary for fast verification
Answer: Load the data directly into BigQuery and apply the ARIMA model type available in BigQuery ML, taking advantage of its seamless integration with Cloud Storage, efficient handling of large datasets, and specialized support for time series forecasting.
### Correct Answer: C **Why this method stands out:** - **Seamless Integration:** BigQuery excels at efficiently managing large datasets, enabling straightforward loading of Avro data and execution of complex queries. - **ARIMA Model Advantage:** Specifically tailored for time series forecasting, the ARIMA model in BigQuery ML is ideal for sales data, adeptly managing seasonality, trends, and data irregularities. - **Quick Iteration:** BigQuery ML facilitates rapid experimentation with various forecasting models and parameters, utilizing SQL for both model creation and evaluation. - **Scalability:** Designed to accommodate growing data volumes, BigQuery ensures your large dataset is processed and analyzed without compromising performance.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
No comments yet.
As a Machine Learning Engineer at a large retailer, you are tasked with developing a model to forecast future sales using 10 years of historical sales data stored in Cloud Storage in Avro format. The company emphasizes rapid experimentation to identify the most effective forecasting model. Key constraints include handling the large dataset efficiently, ensuring the model can capture seasonal trends and irregularities in sales data, and minimizing operational costs. Given these requirements, which approach should you choose to build and train your sales forecast model? Choose the best option.
A
Transform the data into CSV format and utilize the built-in XGBoost algorithm on AI Platform Training, considering its efficiency with large datasets and ability to model complex relationships.
B
Change the data into TFRecords and develop an RNN model using TensorFlow on AI Platform Notebooks, leveraging its capability to capture temporal dependencies in time series data.
C
Load the data directly into BigQuery and apply the ARIMA model type available in BigQuery ML, taking advantage of its seamless integration with Cloud Storage, efficient handling of large datasets, and specialized support for time series forecasting.
D
Convert the data into CSV format and craft a regression model on AutoML Tables, utilizing its automated feature engineering and model selection capabilities for quick experimentation.