Algorithms and hardness for learning linear thresholds from label proportions
R Saket - Advances in Neural Information Processing …, 2022 - proceedings.neurips.cc
We study the learnability of linear threshold functions (LTFs) in the learning from label
proportions (LLP) framework. In this, the feature-vector classifier is learnt from bags of …
proportions (LLP) framework. In this, the feature-vector classifier is learnt from bags of …
Learnability of linear thresholds from label proportions
R Saket - Advances in Neural Information Processing …, 2021 - proceedings.neurips.cc
We study the problem of properly learning linear threshold functions (LTFs) in the learning
from label proportions (LLP) framework. In this, the learning is on a collection of bags of …
from label proportions (LLP) framework. In this, the learning is on a collection of bags of …
Hardness of Learning Boolean Functions from Label Proportions
V Guruswami, R Saket - arXiv preprint arXiv:2403.19401, 2024 - arxiv.org
In recent years the framework of learning from label proportions (LLP) has been gaining
importance in machine learning. In this setting, the training examples are aggregated into …
importance in machine learning. In this setting, the training examples are aggregated into …
[PDF][PDF] Learnability of Linear Thresholds from Label Proportions (Supplemental Appendix)
R Saket - proceedings.neurips.cc
Pr g [pos (〈 z1, g〉)= pos (〈 z2, g〉)]= θ/π,(1) where g∼ N (0, 1) d is a vector with each
coordinate iid standard Gaussian. Note that (i) if〈 z1, z2〉≥ 0, then θ≤ π/2, and therefore …
coordinate iid standard Gaussian. Note that (i) if〈 z1, z2〉≥ 0, then θ≤ π/2, and therefore …