Learn Before
Explain the relationship between beam search approximations and scoring function optimization during inference.
Question: Discuss why approximate search algorithms such as beam search are used during inference to optimize a scoring function like Score_A(S), and explain why these algorithms are not guaranteed to identify the output that maximizes this score.
Sample answer: Beam search is an approximate search algorithm used during inference to find the value of S that optimizes (maximizes) a scoring function, Score_A(S). To manage search complexity, beam search keeps only the top K candidates at each step of the search process. Because it is an approximate method that discards other potential candidates rather than performing an exhaustive search, it is not guaranteed to find the value of S that maximizes Score_A(S).
Key points:
- Beam search is an approximate search algorithm used to find the value of S that maximizes Score_A(S).
- Beam search keeps only the top K candidates during the search process.
- Approximate search algorithms are not guaranteed to find the value of S that maximizes Score_A(S).
Rubric: The response must contain: 1) Identification of beam search as an approximate search algorithm used to optimize/maximize Score_A(S). 2) Explanation that beam search keeps only the top K candidates during the search process. 3) Statement that approximate search algorithms like beam search do not guarantee finding the output that maximizes the scoring function.
0
1
Tags
Machine Learning
Deep Learning
Supervised Learning
Dive into Deep Learning @ D2L
Data Science
Machine Learning Strategy
Related
What does beam search retain at each step of the search process when optimizing a scoring function?
True or False: Beam search is guaranteed to find the output that maximizes the scoring function.
Beam search is an example of an _____ search algorithm that keeps only the top K candidates during the search process.
During beam search, which candidates are retained at each step of the search process?
Beam search is guaranteed to find the output S that maximizes Score_A(S).
Beam search keeps only the top _____ candidates during the search process.
Match each beam search term to its correct description.
Arrange the steps of applying approximate search in a scored inference pipeline.
Why is beam search classified as an 'approximate' search algorithm rather than an exact one?
Beam search is one example of an approximate search algorithm used to optimize a scoring function during inference.
Because beam search is approximate, it is not guaranteed to find the output that _____ Score_A(S).
Match each property of beam search to the consequence it produces.
Order the reasoning steps that explain why beam search may fail to return the globally optimal output.
Explain the relationship between beam search approximations and scoring function optimization during inference.
Diagnosing search limitations when beam search outputs a sub-optimal candidate.
Describe the limitation of using beam search to optimize a scoring function during inference.