{
  "id": "1811.03679",
  "version": "v1",
  "published": "2018-11-08T21:04:00.000Z",
  "updated": "2018-11-08T21:04:00.000Z",
  "title": "Practical Bayesian Learning of Neural Networks via Adaptive Subgradient Methods",
  "authors": [
    "Arnold Salas",
    "Stefan Zohren",
    "Stephen Roberts"
  ],
  "comment": "Manuscript under review by AISTATS 2019",
  "categories": [
    "stat.ML",
    "cs.LG"
  ],
  "abstract": "We introduce a novel framework for the estimation of the posterior distribution of the weights of a neural network, based on a new probabilistic interpretation of adaptive subgradient algorithms such as AdaGrad and Adam. Having a confidence measure of the weights allows several shortcomings of neural networks to be addressed. In particular, the robustness of the network can be improved by performing weight pruning based on signal-to-noise ratios from the weight posterior distribution. Using the MNIST dataset, we demonstrate that the empirical performance of Badam, a particular instance of our framework based on Adam, is competitive in comparison to related Bayesian approaches such as Bayes By Backprop.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2018-11-08T21:04:00.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "neural network",
      "adaptive subgradient methods",
      "practical bayesian learning",
      "weight posterior distribution",
      "novel framework"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}