Uncertainty Decomposition using Multi-objective Model Estimation – We design a simple yet practical method for modeling a real-valued data set. Our proposal is to use Markov Decision Processes (MDPs) for the task of modeling a set of data sets. We show that a Markov Decision Process (MDP) can be trained to model a set of models that are uncertain by using the optimal policies, which is the optimal policy that maximizes the expected utility of each model. We present an optimization strategy that does not require knowledge of the variables that control the policy that maximizes the expected utility of each model. We show that when the variables are unknown, MDPs are flexible and can be used for modelling uncertainty. Our method is not only simple, but also provides the best performance in terms of the optimal policy. The resulting MDP model is a simple, low-latent, yet efficient model of a real-world data set, which is an important data and environment for many real-life applications.
We propose an efficient and flexible variant of Gaussian mixture models that generalizes the linear regression model to the multivariate data. We show that, unlike the linear regression model, the gradient of the covariance matrix, whose function is modeled as the sum of the sum of its Gaussian components, the covariance matrix also matures with Gaussian components, and provides a computationally robust method for the estimation of the covariance matrix. This extension allows us to apply our method to two real-world datasets, representing the physical motions of objects (e.g. human hands and feet) and their visual appearance (e.g. the color of wheels). Experimental results show that our method significantly outperforms the standard method on both tasks, outperforming the traditional one-class classification system on both datasets.
Robust Multidimensional Segmentation based on Edge Prediction
Deep Multi-Objective Goal Modeling
Uncertainty Decomposition using Multi-objective Model Estimation
Mapping Images and Video Summaries to Event-Paths
Stochastic optimization via generative adversarial computingWe propose an efficient and flexible variant of Gaussian mixture models that generalizes the linear regression model to the multivariate data. We show that, unlike the linear regression model, the gradient of the covariance matrix, whose function is modeled as the sum of the sum of its Gaussian components, the covariance matrix also matures with Gaussian components, and provides a computationally robust method for the estimation of the covariance matrix. This extension allows us to apply our method to two real-world datasets, representing the physical motions of objects (e.g. human hands and feet) and their visual appearance (e.g. the color of wheels). Experimental results show that our method significantly outperforms the standard method on both tasks, outperforming the traditional one-class classification system on both datasets.