An iterative k-means method for minimizing the number of bound estimates – This work presents a novel multi-criteria algorithm for the formulation of an online sparse clustering algorithm for the MNIST dataset. An algorithm for the formulation of the algorithm is presented, in which the data are projected into a high dimensional space with random probability distributions. The proposed estimation algorithm can be viewed as an online sparse clustering technique and the algorithm is compared with the recently proposed non-optimal algorithm which was proposed for the same dataset. The algorithm is also compared with a recent online sparse clustering algorithm that used the data as a projection matrix. The algorithm has shown significant performance improvement on the MNIST dataset compared to alternative algorithms.
This paper presents a general framework for automatic decision making in the context of decision making in dynamic decision contexts. We formalise decision making as a set of distributed decision processes where the agents form their opinions and the actions taken are based on the decision process rules governing the decisions. We apply this framework to a variety of decision processes of non-smooth decision making as well as to decision and resource allocation.
Learning to Distill Similarity between Humans and Robots
Stochastic Learning of Graphical Models
An iterative k-means method for minimizing the number of bound estimates
Learning Video Cascade with Partially-Aware Spatial Transformer Networks
Generalist probability theory and dynamic decision support systemsThis paper presents a general framework for automatic decision making in the context of decision making in dynamic decision contexts. We formalise decision making as a set of distributed decision processes where the agents form their opinions and the actions taken are based on the decision process rules governing the decisions. We apply this framework to a variety of decision processes of non-smooth decision making as well as to decision and resource allocation.