Selecting Reliable Levers for a Stagnant Machine Learning Project
Case context: A machine learning engineering team is tasked with raising the performance of a speech recognition model. They need to choose the most reliable development levers to achieve this goal.
Question: According to the principles in the text, what two concrete steps should the team decide to take to reliably improve their model's performance?
Sample answer: The team should decide to train a bigger network and get more data, as these are the two reliable levers for improving algorithm performance.
Key points:
- Decide to train a bigger network to improve performance.
- Decide to get more data to improve performance.
- Recognize these actions as the reliable levers described in the text.
Rubric: Check if the student correctly applies the text's principles by recommending the team train a bigger network and obtain more data.
0
1
Tags
Machine Learning
Deep Learning
Supervised Learning
Dive into Deep Learning @ D2L
Data Science
Machine Learning Strategy
Related
According to Machine Learning Yearning (p. 10), which two strategies are described as reliable ways to improve an algorithm's performance?
True or False: According to Machine Learning Yearning, training a bigger network and getting more data are the ONLY reliable ways to improve an algorithm's performance.
According to Machine Learning Yearning, one of the more reliable improvement levers is to train a bigger _____ and to get more data.
Which two actions does Andrew Ng identify as reliable ways to improve an ML algorithm's performance today?
True or False: Machine Learning Yearning describes training a bigger network and getting more data as reliable ways to improve algorithm performance.
According to Machine Learning Yearning, one reliable way to improve performance is to train a bigger _____ and get more data.
Match each improvement lever from Machine Learning Yearning to its description.
Order the steps for applying the two reliable improvement levers described in Machine Learning Yearning.
What does Andrew Ng mean by calling training a bigger network and getting more data 'reliable' improvement levers?
True or False: According to Machine Learning Yearning, getting more data is an unreliable way to improve an algorithm's performance.
According to Machine Learning Yearning, one reliable way to improve performance is to train a bigger network and get more _____.
Match each bottleneck scenario to the reliable improvement lever from Machine Learning Yearning that most directly addresses it.
Order the reasoning steps that explain why bigger networks and more data are reliable improvement levers per Machine Learning Yearning.
Analyzing the Dual Levers of Modern Algorithm Improvement
Selecting Reliable Levers for a Stagnant Machine Learning Project
Reliable Improvement Levers for Algorithms