Databricks Machine Learning Professional exam dumps

Databricks Machine Learning Professional practice question 256 of 280

Databricks Certified Machine Learning Professional. Professional level, Databricks. Free question with the correct answer and a full explanation.

Databricks Machine Learning Professional Question 256

Select 2

You are tasked with monitoring numerical data for drift in an e-commerce platform's user behavior features. You are considering using either the Jensen-Shannon (JS) divergence or the Kolmogorov-Smirnov (KS) test. Which of the following statements correctly compare these methods for numerical drift detection?

  1. A

    The Kolmogorov-Smirnov test is non-parametric, while the Jensen-Shannon divergence assumes a probabilistic distribution for comparison.

  2. B

    Jensen-Shannon divergence provides a symmetric measure of how two probability distributions differ, while the Kolmogorov-Smirnov test focuses on the maximum difference in cumulative distributions.

  3. C

    The Kolmogorov-Smirnov test requires binning of numerical data before application, while Jensen-Shannon divergence does not.

  4. D

    Jensen-Shannon divergence outputs a specific statistical p-value to determine significance, while the Kolmogorov-Smirnov test does not.

  5. E

    The Kolmogorov-Smirnov test is more suitable for detecting changes in feature distributions over time, while Jensen-Shannon divergence is better for comparing static distributions.

Show answer and explanation

Correct answers: A, B

Explanation

The Kolmogorov-Smirnov (KS) test and Jensen-Shannon (JS) divergence are both used for drift detection, but they differ in approach. The KS test is non-parametric and identifies the maximum difference in cumulative distributions, making it useful for hypothesis testing with a p-value output. JS divergence measures dissimilarity in a symmetric way, assuming probability distributions as input. These distinctions make them suitable for different scenarios depending on the data type and analysis goals.

  • A. Correct.

    Correct. The Kolmogorov-Smirnov test is non-parametric and directly compares cumulative distributions, while JS divergence operates on probability distributions and assumes they are well-defined.

  • B. Correct.

    Correct. JS divergence measures the dissimilarity between two distributions in a symmetric manner, whereas the KS test identifies the largest deviation between cumulative distributions.

  • C. Incorrect.

    Incorrect. The Kolmogorov-Smirnov test does not require binning of numerical data, as it directly measures differences in cumulative distributions.

  • D. Incorrect.

    Incorrect. Jensen-Shannon divergence does not output a p-value; it simply provides a divergence score, while the KS test does output a p-value to indicate statistical significance.

  • E. Incorrect.

    Incorrect. Both the Kolmogorov-Smirnov test and Jensen-Shannon divergence can be used for static or time-series comparisons, but the choice depends on the type of drift and the data characteristics.

Timed practice exam

Take a Databricks Machine Learning Professional practice test under exam conditions

60 questions in 120 minutes, drawn from this bank, with a score report and a per-question review when you finish.

Start timed exam