An Online Convex Optimization Approach for Multi-Relational Time Series Prediction – In this paper, we propose a nonlinear adaptive strategy for non-linear regression using an unsupervised method. Although very useful to model dynamic processes in data analytics, the proposed adaptive strategy is a nonparametric nonparametric regularizer, which is not applicable in the natural data analysis setting where regularity measures are used. We provide an empirical comparison with recent non-stationary regularizers on simulated and real data using simulated and real data sets. The empirical analysis results indicate that while stochastic methods for non-linear regression are effective, the proposed method is not suitable in cases with high non-linearity.

We propose a method for predicting the next stage of an interaction in an evolutionary process that is known to be complex and requires a detailed analysis of the world. This method is a key to a wider understanding of such complex and difficult interaction.

On the Connectivity between Reinforcement Learning and the Game of Go

A deep residual network for event prediction

An Online Convex Optimization Approach for Multi-Relational Time Series Prediction

On the Geometry of Optimal Algorithms for Generalized Support Vector Machines

Learning Multiscale Graph Representations Using Joint Random Field and Graph CutsWe propose a method for predicting the next stage of an interaction in an evolutionary process that is known to be complex and requires a detailed analysis of the world. This method is a key to a wider understanding of such complex and difficult interaction.