Building Predictive Maintenance Models with Images and Explainability

PMLE By islever Editorial Last updated

Serving models. Monitoring, testing, and troubleshooting AI solutions.

Short answer

C Use TensorFlow to create a deep learning-based model, and use Integrated Gradients to explain the model output.

Question

You are building a predictive maintenance model to preemptively detect part defects in bridges. You plan to use high definition images of the bridges as model inputs. You need to explain the output of the model to the relevant stakeholders so they can take appropriate action. How should you build the model?
  1. Use scikit-learn to build a tree-based model, and use SHAP values to explain the model output.
  2. Use scikit-learn to build a tree-based model, and use partial dependence plots (PDP) to explain the model output.
  3. Use TensorFlow to create a deep learning-based model, and use Integrated Gradients to explain the model output. Correct Answer
  4. Use TensorFlow to create a deep learning-based model, and use the sampled Shapley method to explain the model output.

Key Takeaways

The main trap is ignoring the data format; high-definition images necessitate deep learning, not tree-based models, dictating the choice of explanation technique.

Learn how to select appropriate deep learning architectures and explanation methods like Integrated Gradients for image-based predictive maintenance scenarios on PMLE.

Common mistake: Candidates often choose tree-based models (A/B) because they are simpler, failing to recognize that raw image data requires neural networks for effective feature learning.

Expert Analysis

Model Architecture Selection The scenario involves processing high-definition images for defect detection. Image data is high-dimensional and unstructured, requiring feature extraction that tree-based models (like those in scikit-learn) cannot handle directly without extensive manual preprocessing. Deep learning models, specifically Convolutional Neural Networks (CNNs) implemented via frameworks like TensorFlow, are the standard for visual tasks.

Explainability Methodology Stakeholders require model output explanations. For deep learning models, Integrated Gradients is a robust attribution method that calculates the gradient of the output with respect to the input pixels, highlighting which parts of the image contributed to the prediction. This aligns with Google Cloud's Explainable AI guidance for differentiable models.

Why Other Options Fail Options A and B propose tree-based models, which are unsuitable for raw image inputs. Option D suggests the sampled Shapley method; while valid for general features, Integrated Gradients is more efficient and standard for pixel-level explanations in deep learning contexts as per exam doctrine.

Adjacent Decisions & Consequences This question tests the alignment between data type (images) and algorithm complexity. If the input were structured tabular data, a tree-based model might be preferred for simplicity. Choosing the wrong explainability method (e.g., SHAP for deep nets without approximation) leads to computational inefficiency or inaccurate attributions.

Official Reference

https://cloud.google.com/vertex-ai/docs/explainable-ai/overview

This link opens the official vendor documentation in a new tab.

Exam Strategy

Always match the data modality to the model class first. For images, assume deep learning. Then, select an explanation technique native to that architecture (e.g., Integrated Gradients for neural nets) rather than generic feature importance tools designed for tabular data.

Frequently Asked Questions

Why not use SHAP for deep learning models?

While SHAP can be adapted for deep learning, it is computationally expensive. Integrated Gradients is more efficient and theoretically grounded for differentiable models like CNNs.

Can scikit-learn process high-definition images directly?

No. Scikit-learn models expect flattened, numerical feature vectors. Raw images must be preprocessed into features before using tree-based models.

More PMLE FAQ →

Community Votes

C
100%

100% of anonymous learners picked answer C. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 8 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 8

  1. Shark0 👍 4 Answered C
    Given the scenario of using high definition images as inputs for predictive maintenance on bridges, and the need to explain the model output to stakeholders, the most appropriate choice would be: C. Use TensorFlow to create a deep learning-based model, and use Integrated Gradients to explain the model output. Integrated Gradients is a method used to explain the predictions of deep learning models by attributing the contribution of each pixel in the input image to the final prediction. This would provide insights into which parts of the bridge images are most influential in the model's decision-making process, helping stakeholders understand why a particular prediction was made and allowing them to take appropriate action.
  2. dija123 👍 3 Answered C
    Use Integrated Gradients to explain the model output
  3. FireAtMe 👍 2 Answered C
    The question is about image/pixels. So the integrated Gradients is better. Shapley is for input features.
  4. BlehMaks 👍 2 Answered C
    https://cloud.google.com/ai-platform/prediction/docs/ai-explanations/overview#compare-methods
  5. pinimichele01 👍 2 Answered C
    https://cloud.google.com/vertex-ai/docs/explainable-ai/overview
Show 3 more comments
  1. pikachu007 👍 1 Answered C
    Handling image input: Deep learning models excel in processing complex visual data like high-definition images, making them ideal for extracting relevant features from bridge images for defect detection. Explainability with Integrated Gradients: Integrated Gradients is a powerful technique specifically designed to explain the predictions of deep learning models. It attributes model output to specific input features, providing insights into how the model makes decisions. Visualization: Integrated Gradients can generate visual explanations, such as heatmaps, that highlight image regions most influential to predictions, aiding in understanding and trust for stakeholders.
  2. taylz876 👍 1 Answered C
    There are two decisions here: which model to build, and how to explain it. • The model: the inputs are high-definition images of bridges. Raw images are best handled by a deep learning model (a CNN), not a tree-based model. • The explanation: Integrated Gradients is the explainability method for differentiable models like neural networks, and it works well on images with a huge number of input features. It highlights which pixels or regions pushed the prediction toward "defect", so stakeholders can see where the damage is and take action. It is available through Explainable AI in Agent Platform (formerly Vertex AI).
  3. pinimichele01 👍 1
    https://cloud.google.com/vertex-ai/docs/explainable-ai/overview this is right, your is deprecated!

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

Monitoring ML Model Performance with Precision and Recall Sampling

S-grade

Optimizing ML Model Recall for Compliance Risk

S-grade

How to Minimize Development Time for Weekly TensorFlow Retraining in Vertex AI?

S-grade

Building CLV Models with BigQuery ML in Vertex AI Workbench

S-grade

Practice All PMLE Questions

Access 65 questions with complete answers and detailed explanations.

View Full PMLE Practice Test →

← Back to PMLE Study Guide