Rethinking Bi-Level Optimization in Neural Architecture Search: A Gibbs Sampling Perspective.
Chao XueXiaoxing WangJunchi YanYonggang HuXiaokang YangKewei SunPublished in: AAAI (2021)
Keyphrases
- bi level
- neural architecture
- gibbs sampling
- gray scale
- search algorithm
- neural network
- feed forward
- topic models
- approximate inference
- computer vision
- parameter estimation
- expectation maximization
- image compression
- basis functions
- multi layer perceptron
- search space
- em algorithm
- color images
- evolutionary algorithm
- training data
- feature extraction