IBM Planning Analytics V2.0 Developer — Question 15

Someone is training a machine learning system to perform classification of unstructured content. What is the purpose of a validation data set on the training process?

Answer options

• A. Validation data sets are used to fine tune the accuracy of the training model
• B. Validation data sets are only used if test data is unavailable
• C. Validation data sets are exactly the same as test data sets and are used to test the limits of the machine learning model
• D. Validation data sets would never be used in a classification system as only training and test data are required

Correct answer: C

Explanation

The correct answer is C because validation data sets serve to evaluate the performance of the model and are indeed similar to test data sets. Option A is incorrect because while validation data sets help tune the model, they are not solely for accuracy fine-tuning. Option B is misleading as validation data sets are typically used regardless of the availability of test data. Option D is incorrect since validation data is an important part of the model evaluation process in machine learning.