Google Professional Machine Learning Engineer
Get started today
Ultimate access to all questions.
Ultimate access to all questions.
You are a Machine Learning Engineer working on a project that requires deploying models on Google Cloud Platform's Vertex AI. The project's goal is to leverage Explainable AI to understand the most essential features and their influence on the model predictions. The team is considering various model types for deployment. Given the constraints of cost, compliance, and the need for scalability, which of the following models can you apply Vertex Explainable AI to for achieving the project's objectives? (Select three options)
Explanation:
Vertex Explainable AI is designed to provide insights into how models make decisions, particularly for models where understanding the decision-making process is complex. It supports structured data models (like AutoML for classification and regression) and custom-trained models with tabular data and images. Deep Learning models (DNN) and Image Classification models benefit from Explainable AI as they are inherently complex. Decision Trees, however, are inherently interpretable and do not require Explainable AI for understanding. The tool uses methods like sampled Shapley and integrated gradients for feature attributions. For more details, refer to Google's MLOps whitepaper and Vertex AI documentation.