Google Professional Machine Learning Engineer Question 177
Select 2Google Cloud PlatformYour organization is building a machine learning pipeline for processing large-scale streaming data and training a predictive model. The pipeline needs to handle distributed data preprocessing, support scalability, and integrate with TensorFlow for model training. Which of the following tools or frameworks would you choose to meet these requirements?
- A
TensorFlow
- B
PyTorch
- C
Apache Spark
- D
JAX
- E
scikit-learn
Show answer and explanation
Correct answers: A, C
Explanation
To meet the requirements of handling large-scale streaming data and supporting scalability, Apache Spark is the best choice for distributed data preprocessing. TensorFlow is well-suited for training the predictive model, as it integrates seamlessly with Spark for scalable workflows. While other frameworks like PyTorch, JAX, and scikit-learn are useful for specific tasks, they do not fully address the distributed processing and streaming needs in this scenario.
- A. Correct.
TensorFlow is a powerful framework for building and training machine learning models, and it integrates well with scalable distributed processing pipelines.
- B. Incorrect.
PyTorch is also a machine learning framework, but it does not natively provide strong support for distributed data preprocessing or streaming systems like Spark does.
- C. Correct.
Apache Spark is specifically designed for distributed data processing and is well-suited for handling large-scale streaming or batch data before feeding it into a machine learning framework.
- D. Incorrect.
JAX is primarily used for high-performance numerical computing and differentiable programming, but it does not directly support distributed data preprocessing or large-scale streaming pipelines.
- E. Incorrect.
scikit-learn is a great library for implementing traditional machine learning models, but it does not have built-in support for large-scale distributed data processing or integration with streaming pipelines.