T-distributed multi-objective regression with stochastic support vector machines – In this paper we present a method for efficiently performing regression in data-rich, sparse and sparsely represented environments. We show how to combine all the features learnt from different data domains together to perform regression. Our method is inspired by Bayesian process learning, which requires the data to be sampled from unseen sources. We show how to compute a sparse representation of the resulting structure by exploiting sparsity over multiple data domains. To build a non-negative matrix, the non-negative matrix is built into a vector space with a nonzero sum of all the data points. Each sparsity vector can be extracted using a stochastic gradient descent algorithm to form a sparse Euclidean projection. Using a simple but powerful graph embedding technique we show how to use this sparse representation and use it to create a sparse-like embedding matrix. Experimental results on three large datasets with varying sampling rates demonstrate the effectiveness of our approach.
Object segmentation has been extensively used as a tool to train human-machine interfaces to recognize objects and identify them in video. These systems have been shown to produce good results in the context of object classification tasks, but are prone to overfitting when the objects are different than the background. To address this problem, we extend segmentation to a two dimensional space through deep learning to model object semantic and object-specific representations, respectively. The results show that our approach achieves significant improvements in the semantic image segmentation task in terms of accuracy and robustness.
Learning to predict footballs using deep learning
Automatic Dental Bioavailability test using hybrid method
T-distributed multi-objective regression with stochastic support vector machines
Towards the Application of Deep Reinforcement Learning in Wireless LAN Sensor Networks
DenseNet: Generating Multi-Level Neural Networks from End-to-End Instructional VideosObject segmentation has been extensively used as a tool to train human-machine interfaces to recognize objects and identify them in video. These systems have been shown to produce good results in the context of object classification tasks, but are prone to overfitting when the objects are different than the background. To address this problem, we extend segmentation to a two dimensional space through deep learning to model object semantic and object-specific representations, respectively. The results show that our approach achieves significant improvements in the semantic image segmentation task in terms of accuracy and robustness.