arXiv subjects

Aijun An

Publications and source records attributed to Aijun An.

23 records · Page 2Linked to original sources

arXiv2018

Using Neural Network for Identifying Clickbaits in Online News Media

Online news media sometimes use misleading headlines to lure users to open the news article. These catchy headlines that attract users but disappointed them at the end, are called Clickbaits. Because of the importance of automatic clickbait detection in online medias, lots of machine learning methods were proposed and employed to find the clickbait headlines. In this research, a model using deep learning methods is proposed to find the clickbaits in Clickbait Challenge 2017's dataset. The proposed model gained the first rank in the Clickbait Challenge 2017 in terms of Mean Squared Error. Also, data analytics and visualization techniques are employed to explore and discover the provided dataset to get more insight from the data.

cs.CL↗

arXiv2017

Spontaneous Symmetry Breaking in Neural Networks

We propose a framework to understand the unprecedented performance and robustness of deep neural networks using field theory. Correlations between the weights within the same layer can be described by symmetries in that layer, and networks generalize better if such symmetries are broken to reduce the redundancies of the weights. Using a two parameter field theory, we find that the network can break such symmetries itself towards the end of training in a process commonly known in physics as spontaneous symmetry breaking. This corresponds to a network generalizing itself without any user input layers to break the symmetry, but by communication with adjacent layers. In the layer decoupling limit applicable to residual networks (He et al., 2015), we show that the remnant symmetries that survive the non-linear layers are spontaneously broken. The Lagrangian for the non-linear and weight layers together has striking similarities with the one in quantum field theory of a scalar. Using results from quantum field theory we show that our framework is able to explain many experimentally observed phenomena,such as training on random labels with zero error (Zhang et al., 2017), the information bottleneck, the phase transition out of it and gradient variance explosion (Shwartz-Ziv & Tishby, 2017), shattered gradients (Balduzzi et al., 2017), and many more.

stat.CO↗

arXiv2017

Optimization assisted MCMC

Markov Chain Monte Carlo (MCMC) sampling methods are widely used but often encounter either slow convergence or biased sampling when applied to multimodal high dimensional distributions. In this paper, we present a general framework of improving classical MCMC samplers by employing a global optimization method. The global optimization method first reduces a high dimensional search to an one dimensional geodesic to find a starting point close to a local mode. The search is accelerated and completed by using a local search method such as BFGS. We modify the target distribution by extracting a local Gaussian distribution aound the found mode. The process is repeated to find all the modes during sampling on the fly. We integrate the optimization algorithm into the Wormhole Hamiltonian Monte Carlo (WHMC) method. Experimental results show that, when applied to high dimensional, multimodal Gaussian mixture models and the network sensor localization problem, the proposed method achieves much faster convergence, with relative error from the mean improved by about an order of magnitude than WHMC in some cases.

stat.CO↗

arXiv2016

Authority-based Team Discovery in Social Networks

Given a social network of experts, we address the problem of discovering a team of experts that collectively holds a set of skills required to complete a given project. Most prior work ranks possible solutions by communication cost, represented by edge weights in the expert network. Our contribution is to take experts authority into account, represented by node weights. We formulate several problems that combine communication cost and authority, we prove that they are NP-hard, and we propose and experimentally evaluate greedy algorithms to solve them.

cs.SI↗

arXiv2015

Geodesic and Contour Optimization Using Conformal Mapping

We propose a novel optimization algorithm for continuous functions using geodesics and contours under conformal mapping.The algorithm can find multiple optima by first following a geodesic curve to a local optimum then traveling to the next search area by following a contour curve. To improve the efficiency, Newton-Raphson algorithm is also employed in local search steps. A proposed jumping mechanism based on realized geodesics enables the algorithm to jump to a nearby region and consequently avoid trapping in local optima. Conformal mapping is used to resolve numerical instability associated with solving the classical geodesic equations. Geodesic flows under conformal mapping are constructed numerically by using local quadratic approximation. The parameters in the algorithm are adaptively chosen to reflect local geometric features of the objective function. Comparisons with many commonly used optimization algorithms including gradient, trust region, genetic algorithm and global search methods have shown that the proposed algorithm outperforms most widely used methods in almost all test cases with only a couple of exceptions.

stat.CO↗