Answer-first summary for fast verification
Answer: Use the TFX ModelValidator tools to specify performance metrics for production readiness.
The correct answer is A. The Evaluator component in TensorFlow Extended (TFX) allows you to evaluate model performance on various subsets of data, ensuring that the model meets the specific metrics required for production readiness. This approach is streamlined and reliable as it provides deep analysis on training results and helps you understand model behavior on different segments of your data. Using k-fold cross-validation (Option B) is generally not suitable for time series data due to potential data leakage issues. Option C, while relevant for time-sensitive data, does not explicitly address the need to track performance across multiple data subsets. Option D focuses on a singular metric (AUC ROC) using the entire dataset, which does not align with the requirement to track performance on specific subsets before production.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
You are working as a data scientist for a global footwear retailer, responsible for predicting when an item will be out of stock based on historical inventory data. The demand for footwear is highly dynamic as it is influenced by numerous external factors such as fashion trends, seasonality, and marketing campaigns. Your goal is to develop and deploy models trained on all available data while specifically tracking performance on different subsets of data to ensure reliability before pushing to production. What is the most streamlined and reliable way to perform this validation?
A
Use the TFX ModelValidator tools to specify performance metrics for production readiness.
B
Use k-fold cross-validation as a validation strategy to ensure that your model is ready for production.
C
Use the last relevant week of data as a validation set to ensure that your model is performing accurately on current data.
D
Use the entire dataset and treat the area under the receiver operating characteristics curve (AUC ROC) as the main metric.
No comments yet.