Efficiently testing local optimality and escaping saddles for ReLU networks
September 28, 2018 Β· Declared Dead Β· π International Conference on Learning Representations
"No code URL or promise found in abstract"
Evidence collected by the PWNC Scanner
Authors
Chulhee Yun, Suvrit Sra, Ali Jadbabaie
arXiv ID
1809.10858
Category
math.OC: Optimization & Control
Cross-listed
cs.LG,
stat.ML
Citations
10
Venue
International Conference on Learning Representations
Last Checked
5 months ago
Abstract
We provide a theoretical algorithm for checking local optimality and escaping saddles at nondifferentiable points of empirical risks of two-layer ReLU networks. Our algorithm receives any parameter value and returns: local minimum, second-order stationary point, or a strict descent direction. The presence of $M$ data points on the nondifferentiability of the ReLU divides the parameter space into at most $2^M$ regions, which makes analysis difficult. By exploiting polyhedral geometry, we reduce the total computation down to one convex quadratic program (QP) for each hidden node, $O(M)$ (in)equality tests, and one (or a few) nonconvex QP. For the last QP, we show that our specific problem can be solved efficiently, in spite of nonconvexity. In the benign case, we solve one equality constrained QP, and we prove that projected gradient descent solves it exponentially fast. In the bad case, we have to solve a few more inequality constrained QPs, but we prove that the time complexity is exponential only in the number of inequality constraints. Our experiments show that either benign case or bad case with very few inequality constraints occurs, implying that our algorithm is efficient in most cases.
Community Contributions
Found the code? Know the venue? Think something is wrong? Let us know!
π Similar Papers
In the same crypt β Optimization & Control
R.I.P.
π»
Ghosted
R.I.P.
π»
Ghosted
Local SGD Converges Fast and Communicates Little
R.I.P.
π»
Ghosted
On Lazy Training in Differentiable Programming
π
π
The Cartographer
A Review on Bilevel Optimization: From Classical to Evolutionary Approaches and Applications
R.I.P.
π»
Ghosted
Learned Primal-dual Reconstruction
R.I.P.
π»
Ghosted
On the Global Convergence of Gradient Descent for Over-parameterized Models using Optimal Transport
Died the same way β π» Ghosted
R.I.P.
π»
Ghosted
Federated Learning: Strategies for Improving Communication Efficiency
R.I.P.
π»
Ghosted
In-Datacenter Performance Analysis of a Tensor Processing Unit
R.I.P.
π»
Ghosted
Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning
R.I.P.
π»
Ghosted