AI Engineering Degree Practice Exam 2025 - Free AI Engineering Practice Questions and Study Guide

The scenario where a model exhibits high training accuracy yet low out-of-sample accuracy is often indicative of overfitting. Overfitting occurs when a model learns to memorize the training data rather than generalizing from it. When a model is trained on insufficient data, it tends to capture the noise and specific patterns of that dataset, which does not reflect the broader distribution of data it will encounter in real-world applications.

This leads to the model performing exceptionally well on the training dataset since it has essentially 'memorized' it, but when faced with new, unseen data (the out-of-sample data), the model struggles to make accurate predictions because it cannot generalize beyond the patterns it learned. Thus, the high training accuracy does not translate to effectiveness in practical applications, resulting in low out-of-sample accuracy.

Other options do not provide the same context in relation to the training and out-of-sample accuracy. For instance, optimizing a model for speed rather than accuracy may impact performance metrics but does not inherently link to the dichotomy between training and out-of-sample performance. Similarly, while including too many variables or using default settings could lead to various modeling problems, they do not specifically address the issue of overfitting, which is