Test Driven Development for Neural Networks, Part II — AB Testing

  1. Simple single-layer linear
  2. Simple convolutional network
  3. Deep convolutional network
  1. Stochastic Gradient Descent
  2. Conjugate Gradient Descent
  3. L-BFGS
  4. OWL-QN
  1. MNIST — A small gray scale image tile classification data set which is easy to achieve ~91% on. Gray scale 28x28 images for each 0–9 digit.
  2. CIFAR — Another, much more difficult small image tile classification data set. Color 32x32 images in 10 categories.
  1. Classification — Learn to classify the images
  2. Representation — Derive a common decoding network and input values which reconstruct the data set
  3. Denoising Autoencoder (in development)
  4. Generative Adversarial Networks (in development)

--

--

--

Big Data Engineer and Artificial Intelligence Researcher

Love podcasts or audiobooks? Learn on the go with our new app.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Andrew Charneski

Andrew Charneski

Big Data Engineer and Artificial Intelligence Researcher

More from Medium

Map Side Join in HIVE

Prim’s Algorithm

Graph Introduction

String view in a Nutshell