You have built a machine model to predict whether a bank customer is going to default on a loan. You want to use Local Interpretable Model-Agnostic Explanations (LIME) to understand a specific prediction. What is the key idea behind LIME?
Correct Answer: A
Detailed Answer in Step-by-Step Solution: Objective: Define LIME’s core concept. Understand LIME: Explains individual predictions with local surrogate models. Evaluate Options: A: Complex global, simple local—Correct LIME principle. B: Agnosticism—True but not the key idea. C: Global/local similarity—False. D: Local vs. global agnosticism—Incorrect distinction. Reasoning: A captures LIME’s local approximation focus. Conclusion: A is correct. OCI documentation notes: “LIME (A) explains predictions by approximating complex global models with simpler local surrogate models around specific instances.” B, C, and D misalign—only A reflects LIME’s foundational idea per OCI’s interpretability tools. 1: Oracle Cloud Infrastructure Data Science Documentation, "Model Interpretability - LIME".
Question 52
A bike sharing platform has collected user commute data for the past 3 years. For increasing profitability and making useful inferences, a machine learning model needs to be built from the accumulated data. Which of the following options has the correct order of the required machine learning tasks for building a model?
Correct Answer: C
Detailed Answer in Step-by-Step Solution: Data Access: The first step in any machine learning workflow is accessing the raw data. This involves retrieving the user commute data collected over the past 3 years from the bike-sharing platform’s storage system. Data Exploration: Once data is accessed, it’s explored to understand its structure, quality, and patterns (e.g., missing values, distributions). This step helps identify what preprocessing is needed. Feature Engineering: After understanding the data, features are created or transformed (e.g., commute duration, time of day) to improve model performance. This step precedes feature exploration because you need engineered features to analyze further. Feature Exploration: This involves analyzing the engineered features (e.g., correlation analysis, importance ranking) to refine them or select the most relevant ones for modeling. Modeling: Finally, the prepared data and features are used to train and evaluate a machine learning model. Option C (Data Access, Data Exploration, Feature Engineering, Feature Exploration, Modeling) follows this logical sequence, aligning with standard ML workflows. The correct order reflects the machine learning lifecycle as outlined in Oracle’s OCI Data Science documentation. Data Access is the initial step to retrieve data, followed by Data Exploration to assess it (e.g., using OCI Data Science Notebook Sessions with tools like pandas). Feature Engineering transforms raw data into meaningful inputs, followed by Feature Exploration to analyze feature importance (e.g., using ADS SDK’s correlation tools). Modeling is the final step where the model is built and trained. This sequence is consistent with Oracle’s recommended practices for building ML models in OCI Data Science (Oracle Cloud Infrastructure Data Science Service Documentation, "Machine Learning Lifecycle").
