csc 591-024, (8290)
csc 791-024, (8291)
fall 2024, special topics in computer science
Tim Menzies, timm@ieee.org, com sci, nc state

home :: timetable :: syllabus :: groups :: moodle :: license

HW4 : Concepts

  1. What is the standard line on “how much data is enough?”
    • From Peter Norvig
    • From regression theory
    • From semi-supervised learning
  2. Describe each of the following. What are their implications for human decision making
    • Streaming over zero-diversity data
    • STM, LTM
    • Shrikanth’s early bird effect
    • two results (Valerdi’s work; repertory grids) commenting on the rate at which we can extract considered opinions from humans
  3. In English, describe the following math results and their implications for data mining
    • chessboard model
    • probable correctness theory.
  4. Few-shot learning (FSL):
    • describe it (use this as a guide: https://www.promptingguide.ai/techniques/fewshot)
    • In what sense does FSL mean we can look at fewer examples?
    • In what sense does FSL require many, many examples
  5. Describe an active learning cycle. include the words acquisition function, exploit, explore, warm/code start
  6. Acquisition functions. Distinguish and define the following terms.  diversity, perversity, (population|surrogate|pool|stream)-based
  7. What is PCA? (Us this as a guide: https://en.wikipedia.org/wiki/Principal_component_analysis)
    • How is ezr’s recursive use of “twoFar” an analog of PCA?
    • Recursive twoFar generates a tree of clusters. What does ezr’s leaf function do? Assuming balances cluster tree are already built, in “big O” notation, what is leaf’s runtime?
    • The ezr functions half and cluster sort the sort flag. What is the impact on #evalautions oil sortp=True
    • (HARD): Currently, in 24Aug12, ezr’s half (called from brach) requires two evals per depth of tree. Can you reduce that to one (for all levels except the top one)
  8. Describe, in Engkish, the k-means algorithm (hint: see “kmeans” in https://github.com/timm/noml/blob/main/src/mink.py)
    • How is the same/different as recursive twoFar clustering?
  9. What is “discretization?” (hint: see https://www.blog.trainindata.com/data-discretization-in-machine-learning/)
    • What is gaussian discretization? Hint: unsuper, the “bin” function in https://github.com/timm/noml/blob/main/src/unsuper.py
    • How should the gaussian discretizer handle symbolic columns (hint:trick question).
    • How does unsuper divide numeric columns (hint: see COLS’bins function and its use of i.bin)
    • Why does unsuper need to merge (some) of the bins it generated ?
    • How does unsuper’s merges function combine tow adjacent bins?