Online Voting for Social Good – We propose a multi-class framework for automatic voting with the aim of improving the quality of the quality of voting among both voters and their voting intention. Multi-class voting (ML) in particular is a form of voting in which voters vote in a single class instead of a different class. However, voting is inherently a subjective process which is subjective and requires different criteria. We propose an algorithm that optimally selects a class of voters in terms of its voting intention. The algorithm is evaluated against several existing voting algorithms in the literature, and results show that it consistently performs favorably when compared with several existing voting algorithms. As a complementary feature of ML, a new class of non-local autoregressive voting algorithms based on a hybrid autoregressive voting process are presented. Experimental results on two different datasets from the literature were used to validate the proposed algorithm. The algorithms show that the proposed algorithm is significantly faster than the existing voting algorithm. More importantly, we show that this method can be applied to other voting systems such as the U-S.A.

We propose a novel distributed optimization method for machine learning. Our goal is to use our method as well as the other popular techniques in learning to map images to objects by a large distance metric to achieve an improved prediction. Our approach aims to train a neural net to predict images according to a few parameters that are relevant to each input image. We show how to use our method to map images to objects using an online model trained on a small set of images. Experimental results were performed on synthetic and real datasets to compare the performance of the proposed method. The results show that our approach provides a better learning rate than conventional supervised learning by the same distance metric.

Lip Localization via Semi-Local Kernels

Towards Optimal Cooperative and Efficient Hardware Implementations

Online Voting for Social Good

Randomized Methods for Online and Stochastic Link Prediction

A Novel Fuzzy Model for Indoor Localization and LocalizationWe propose a novel distributed optimization method for machine learning. Our goal is to use our method as well as the other popular techniques in learning to map images to objects by a large distance metric to achieve an improved prediction. Our approach aims to train a neural net to predict images according to a few parameters that are relevant to each input image. We show how to use our method to map images to objects using an online model trained on a small set of images. Experimental results were performed on synthetic and real datasets to compare the performance of the proposed method. The results show that our approach provides a better learning rate than conventional supervised learning by the same distance metric.