A priori estimates for classification problems using neural networks
arXiv:2009.13500v18 citations
Originality Synthesis-oriented
AI Analysis
This work provides theoretical guarantees for neural network classification, which is incremental as it builds on existing complexity and approximation methods.
The paper tackles the problem of obtaining a priori error estimates for classification using neural networks by applying Rademacher complexity and approximation theorems to regularized loss functionals, resulting in theoretical bounds on error without specifying concrete numbers.
We consider binary and multi-class classification problems using hypothesis classes of neural networks. For a given hypothesis class, we use Rademacher complexity estimates and direct approximation theorems to obtain a priori error estimates for regularized loss functionals.