Google

An error table for the cat image detector can put two data distributions on the x-axis and three error types on the y-axis: human-level error, error on examples the algorithm has trained on, and error on examples the algorithm has not trained on. Filling additional entries may sometimes give additional insight about what the algorithm is doing on the two data distributions.

Error Table Across Two Data Distributions and Three Error Types

Machine Learning Yearning, a free ebook from Andrew Ng, teaches you how to structure Machine Learning projects.
https://www.deeplearning.ai/machine-learning-yearning/

Note: The content of the book is aligned with the Coursera Deeplearning.ai specialization.  https://www.deeplearning.ai/deep-learning-specialization/ 

Machine Learning Yearning (Deeplearning.ai)

In the error table for the cat image detector, what is placed along the x-axis?

The error table for the cat image detector places three data distributions along the x-axis.

The error table places two data _____ on the x-axis and three error types on the y-axis.

Match each axis or row label in the error table to its role in the cat image detector example.

Order the steps for constructing the error table described in Machine Learning Yearning.

What does Andrew Ng say is a benefit of filling in additional entries in the error table?

Andrew Ng states that drawing errors as table entries makes it easier to understand how different error types relate to each other.

One y-axis row in the error table is 'error on examples the algorithm has _____ on.'

Match each error table comparison to the diagnostic insight it reveals about the algorithm.

Order the reasoning steps for diagnosing algorithm behavior using the completed error table.

Question:
Describe the structure of the error table proposed by Andrew Ng for evaluating a model like the cat image detector, including the axes and their respective components. Additionally, explain why an engineer might fill in the additional entries in this table.

Sample answer:
The error table is structured with data distributions on the x-axis and error types on the y-axis. For the cat image detector, the x-axis contains two different data distributions. The y-axis contains three specific error types: human-level error, error on trained examples, and error on untrained examples. Filling in the additional entries in this table is valuable because it can provide deeper diagnostic insight into how the algorithm behaves differently across the two data distributions.

Key points:
- The x-axis represents two different data distributions.
- The y-axis represents human-level error, error on trained examples, and error on untrained examples.
- Filling additional entries provides diagnostic insight into the algorithm's behavior across distributions.

Rubric:
A strong answer should accurately identify the components of the x and y axes (two distributions and three specific error types) and mention that additional entries provide insights into the algorithm's performance across distributions.

Explain the structure and value of the error table in evaluating model performance.

Case context:
An ML team has trained a classifier on high-resolution web images (Distribution A) but plans to deploy it on low-resolution mobile uploads (Distribution B). They want to systematically understand how different types of errors relate to each other using an error table.

Question:
Based on Andrew Ng's framework, how should the team structure their error table to evaluate the classifier, and what specific rows and columns should they include to gain insight into the algorithm's behavior?

Sample answer:
The team should structure the table by placing the two data distributions (Distribution A and Distribution B) on the x-axis. On the y-axis, they should include three rows representing the error types: human-level error, the algorithm's error on examples it has trained on, and the algorithm's error on examples it has not trained on. Filling in this grid will help them compare the errors and diagnose what the algorithm is doing on the different distributions.

Key points:
- Place the two data distributions on the x-axis.
- Place the three error types on the y-axis.
- The error types are human-level error, error on trained examples, and error on untrained examples.

Rubric:
The answer must correctly place the two distributions on the x-axis and the three specific error types on the y-axis, noting that this structure helps diagnose the relationship between different error types.

Structuring diagnostic analysis for a new classifier across distinct data sources.

Question:
According to the structure of the error table used for the cat image detector, what are the three specific types of error placed on the y-axis?

Sample answer:
The three types of error on the y-axis are human-level error, error on examples the algorithm has trained on, and error on examples the algorithm has not trained on.

Key points:
- Human-level error
- Error on examples the algorithm has trained on
- Error on examples the algorithm has not trained on

Rubric:
The response must list all three exact error types mentioned in the text to receive full credit.

Learn Before

Related