Teaching and compressing for low VC-dimension

@article{Moran2015TeachingAC,
  title={Teaching and compressing for low VC-dimension},
  author={Shay Moran and A. Shpilka and A. Wigderson and A. Yehudayoff},
  journal={ArXiv},
  year={2015},
  volume={abs/1502.06187}
}
  • Shay Moran, A. Shpilka, +1 author A. Yehudayoff
  • Published 2015
  • Computer Science, Mathematics
  • ArXiv
  • In this work we study the quantitative relation between VC-dimension and two other basic parameters related to learning and teaching. Namely, the quality of sample compression schemes and of teaching sets for classes of low VC-dimension. Let $C$ be a binary concept class of size $m$ and VC-dimension $d$. Prior to this work, the best known upper bounds for both parameters were $\log(m)$, while the best lower bounds are linear in $d$. We present significantly better upper bounds on both as… CONTINUE READING
    13 Citations
    Compressing and Teaching for Low VC-Dimension
    • 16
    Labeled Compression Schemes for Extremal Classes
    • 14
    • PDF
    Sample compression schemes for VC classes
    • 37
    • PDF
    Open Problem: Recursive Teaching Dimension Versus VC Dimension
    • 14
    • PDF
    Agnostic Sample Compression for Linear Regression
    Hitting Set in hypergraphs of low VC-dimension
    • 8
    • PDF
    Proper PAC learning is compressing
    • 4
    Optimal Collusion-Free Teaching
    • 3
    • Highly Influenced
    • PDF
    Sample Compression for Real-Valued Learners
    • 3
    • PDF

    References

    SHOWING 1-10 OF 52 REFERENCES
    Recursive Teaching Dimension, Learning Complexity, and Maximum Classes
    • 27
    • Highly Influential
    • PDF
    Sample compression schemes for VC classes
    • 11
    A Note on Teaching for VC Classes
    • 3
    • Highly Influential
    On the complexity of teaching
    • 135
    • Highly Influential
    Sample compression, learnability, and the Vapnik-Chervonenkis dimension
    • 83
    • Highly Influential
    • PDF
    Space-bounded learning and the Vapnik-Chervonenkis dimension
    • S. Floyd
    • Mathematics, Computer Science
    • COLT '89
    • 1989
    • 57
    • Highly Influential
    A Geometric Approach to Sample Compression
    • 33
    • PDF
    Learning Binary Relations and Total Orders
    • 45