An AI practitioner has completed model training but accuracy has not reached the target. The practitioner wants to improve model performance by adjusting settings such as the learning rate and batch size. Which process does this describe?

1 / 1
Select an answer
CorrectD

Explanation

Question Overview

Identify the process of adjusting training settings to improve performance.

Requirements to satisfy
  • 1「adjusting settings such as the learning rate and batch size」Adjusting hyperparameters that determine how training proceeds
  • 2「improve model performance」Searching for optimal values to improve performance = hyperparameter tuning
Per-option explanation
AIncorrect

Feature engineering

This is incorrect. Feature engineering is the work of creating effective input variables from raw data for prediction. It is a separate process from adjusting settings such as the learning rate.

BIncorrect

Data collection

This is incorrect. Data collection is the phase of gathering raw data for training. It is a separate process from adjusting settings such as the learning rate.

CIncorrect

Deployment

This is incorrect. Deployment is the phase of releasing the completed model to production. It is a separate process from adjusting settings such as the learning rate.

DCorrect

Hyperparameter tuning

This is correct. Hyperparameter tuning is the process of adjusting hyperparameters — the settings that determine how training proceeds, such as the learning rate, batch size, and number of epochs — to improve model performance.

Key Takeaway

Hyperparameter tuning in context:
- Adjust settings that determine how training proceeds — such as learning rate, batch size, and number of epochs — to improve model performance.
- A process of repeated trials during and after training to find optimal values.
Feature engineering (creating input variables), data collection, and deployment are different phases and purposes.

Related Links