Corpus ID: 15105362

Learning Halfspaces and Neural Networks with Random Initialization

@article{Zhang2015LearningHA,
  title={Learning Halfspaces and Neural Networks with Random Initialization},
  author={Yuchen Zhang and J. Lee and M. Wainwright and Michael I. Jordan},
  journal={ArXiv},
  year={2015},
  volume={abs/1511.07948}
}
We study non-convex empirical risk minimization for learning halfspaces and neural networks. For loss functions that are $L$-Lipschitz continuous, we present algorithms to learn halfspaces and multi-layer neural networks that achieve arbitrarily small excess risk $\epsilon>0$. The time complexity is polynomial in the input dimension $d$ and the sample size $n$, but exponential in the quantity $(L/\epsilon^2)\log(L/\epsilon)$. These algorithms run multiple rounds of random initialization… Expand
Reliably Learning the ReLU in Polynomial Time
Eigenvalue Decay Implies Polynomial-Time Learnability for Neural Networks
On the Quality of the Initial Basin in Overspecified Neural Networks
How Many Samples are Needed to Learn a Convolutional Neural Network?
Gradient Descent Learns One-hidden-layer CNN: Don't be Afraid of Spurious Local Minima
Distribution-Specific Hardness of Learning Neural Networks
  • O. Shamir
  • Computer Science, Mathematics
  • J. Mach. Learn. Res.
  • 2018
Convergence Analysis of Two-layer Neural Networks with ReLU Activation
SGD Learns the Conjugate Kernel Class of the Network
How Many Samples are Needed to Estimate a Convolutional Neural Network?
...
1
2
3
4
...

References

SHOWING 1-10 OF 40 REFERENCES
Learning Kernel-Based Halfspaces with the 0-1 Loss
Efficient Learning of Linear Separators under Bounded Noise
Agnostically learning halfspaces
Efficient Learning of Linear Perceptrons
Generalization Bounds for Neural Networks through Tensor Factorization
Hardness of Learning Halfspaces with Noise
  • V. Guruswami, P. Raghavendra
  • Mathematics, Computer Science
  • 2006 47th Annual IEEE Symposium on Foundations of Computer Science (FOCS'06)
  • 2006
Learning Halfspaces with Malicious Noise
Learning Halfspaces with the Zero-One Loss: Time-Accuracy Tradeoffs
...
1
2
3
4
...