Google Professional-Machine-Learning-Engineer Exam : Google Professional Machine Learning Engineer

You work for a large bank that serves customers through an application hosted in Google Cloud that is running in the US and Singapore. You have developed a PyTorch model to classify transactions as potentially fraudulent or not. The model is a three-layer perceptron that uses both numerical and categorical features as input, and hashing happens within the model.
You deployed the model to the us-central1 region on nl-highcpu-16 machines, and predictions are served in real time. The model's current median response latency is 40 ms. You want to reduce latency, especially in Singapore, where some customers are experiencing the longest delays.
What should you do?

• A. Attach an NVIDIA T4 GPU to the machines being used for online inference.
• B. Change the machines being used for online inference to nl-highcpu-32.
• C. Create another Vertex AI endpoint in the asia-southeast1 region, and allow the application to choose the appropriate endpoint.
• D. Deploy the model to Vertex AI private endpoints in the us-central1 and asia-southeast1 regions, and allow the application to choose the appropriate endpoint.

You are developing a model to identify traffic signs in images extracted from videos taken from the dashboard of a vehicle. You have a dataset of 100,000 images that were cropped to show one out of ten different traffic signs. The images have been labeled accordingly for model training, and are stored in a Cloud Storage bucket. You need to be able to tune the model during each training run. How should you train the model?

• A. Develop the model training code for image classification, and train a model by using Vertex AI custom training.
• B. Train a model for image classification by using Vertex AI AutoML.
• C. Develop the model training code for object detection, and train a model by using Vertex AI custom training.
• D. Train a model for object detection by using Vertex AI AutoML.

You are building an internal assistant that answers employee questions about a 12,000-page HR policy corpus. The corpus is updated weekly, and answers must cite the specific policy section.
Hallucinated answers have been reported during testing. You need to improve factual accuracy while keeping the content current. What should you do?

• A. Chunk and embed the corpus into Vector Search, retrieve the top-k relevant chunks at query time, and pass them to Gemini as grounding context with their source identifiers.
• B. Place the entire corpus in the system instruction of every request, and instruct the model to quote the relevant section.
• C. Increase the model temperature to 0 and instruct the model in the system prompt to answer only from HR policies.
• D. Fine-tune Gemini on the policy corpus each week so the model memorizes the latest policies.

You are tuning a gradient-boosted model with eight hyperparameters. Your previous approach used grid search, which consumed the entire monthly compute budget without converging on a strong configuration. You need an approach that finds a good configuration with fewer trials and automatically abandons unpromising runs. What should you do?

• A. Use Agent Platform hyperparameter tuning with Bayesian optimization and early stopping enabled.
• B. Manually adjust hyperparameters based on the results of each completed run.
• C. Fix six of the hyperparameters at their default values, and grid search the remaining two.
• D. Switch to random search with the same number of trials as the grid search.

Your team has been tasked with creating an ML solution in Google Cloud to classify support requests for one of your platforms. You analyzed the requirements and decided to use TensorFlow to build the classifier so that you have full control of the model's code, serving, and deployment. You will use Kubeflow pipelines for the ML platform. To save time, you want to build on existing resources and use managed services instead of building a completely new model.
How should you build the classifier?

• A. Use AutoML Natural Language to build the support requests classifier.
• B. Use an established text classification model on AI Platform to perform transfer learning.
• C. Use the Natural Language API to classify support requests.
• D. Use an established text classification model on AI Platform as-is to classify support requests.