[link]
# Generative Adversarial Nets ## Introduction * The paper proposes an adversarial approach for estimating generative models where one model (generative model) tries to learn a data distribution and another model (discriminative model) tries to distinguish between samples from the generative model and original data distribution. * [Link to the paper](https://arxiv.org/abs/1406.2661) ## Adversarial Net * Two models  Generative Model(*G*) and Discriminative Model(*D*) * Both are multilayer perceptrons. * *G* takes as input a noise variable *z* and outputs data sample *x(=G(z))*. * *D* takes as input a data sample *x* and predicts whether it came from true data or from *G*. * *G* tries to minimise *log(1D(G(z)))* while *D* tries to maximise the probability of correct classification. * Think of it as a minimax game between 2 players and the global optimum would be when *G* generates perfect samples and *D* can not distinguish between the samples (thereby always returning 0.5 as the probability of sample coming from true data). * Alternate between *k* steps of training *D* and 1 step of training *G* so that *D* is maintained near its optimal solution. * When starting training, the loss *log(1D(G(z)))* would saturate as *G* would be weak. Instead maximise *log(D(G(z)))* * The paper contains the theoretical proof for global optimum of the minimax game. ## Experiments * Datasets * MNIST, Toronto Face Database, CIFAR10 * Generator model uses RELU and sigmoid activations. * Discriminator model uses maxout and dropout. * Evaluation Metric * Fit Gaussian Parzen window to samples obtained from *G* and compare loglikelihood. ## Strengths * Computational advantages * Backprop is sufficient for training with no need for Markov chains or performing inference. * A variety of functions can be used in the model. * Since *G* is trained only using the gradients from *D*, fewer chances of directly copying features from the true data. * Can represent sharp (even degenerate) distributions. ## Weakness * *D* must be well synchronised with *G*. * While *G* may learn to sample data points that are indistinguishable from true data, no explicit representation can be obtained. ## Possible Extensions * Conditional generative models. * Inference network to predict *z* given *x*. * Implement a stochastic extension of the deterministic [MultiPrediction Deep Boltzmann Machines](https://papers.nips.cc/paper/5024multipredictiondeepboltzmannmachines.pdf) * Using discriminator net or inference net for feature selection. * Accelerating training by ensuring better coordination between *G* and *D* or by determining better distributions to sample *z* from during training.
Your comment:
