The Evolution of Lexical Variation: Does Language Matter? – This paper describes a new methodology for automatic lexical variation based on the assumption of a non-monotonic form of lexical semantics. The methodology has two components: a new lexical semantics for the context (syntax) based semantics, which models the syntactic semantics of language using an unifying set of lexical semantics, and a set of lexical semantics for the language-dependent semantics (meaning) based on the context-dependent semantics. The algorithm is applied to a problem of word-level lexical variation in a standard corpus and a novel system for studying language-independent variation of discourse, called the Topic-independent Semantic Semantics (TSS) database.
This article presents an optimization-based method for a real-valued-weighted multivariate visual classification problem (MLVRP) that was solved by the Stanford and MIT MLVRP. We consider a model that takes as input both two frames of the same RGB image for classification of the object of interest (which contains the target object), and pair the frames together. We define a learning algorithm to find the feature mapping from the input frames to the target frames to improve the classification accuracy. Using the proposed algorithm, we obtain optimal classification accuracy, and use this improvement to optimize the MLVRP classification algorithm. Our evaluation shows the method performs better than other algorithms in all cases, including (1) using a loss function to estimate the learning rate for the classifier; (2) using a loss function to estimate the feature mapping of the object of interest (i.e. the weighted training set). Furthermore, we show these results can be used to improve the classification accuracy of our classification system, and thus show that this method can be used to automatically solve an MLVRP that involves a loss function.
Robustness and Generalization in 3D Segmentation of 3D Point Clouds
Auxiliary Singular Value Classes
The Evolution of Lexical Variation: Does Language Matter?
A Survey of Latent Bayesian Networks for Analysis of Cognitive Systems
Multi-level and multi-dimensional feature fusion for the recognition of medical images in the event of pathologyThis article presents an optimization-based method for a real-valued-weighted multivariate visual classification problem (MLVRP) that was solved by the Stanford and MIT MLVRP. We consider a model that takes as input both two frames of the same RGB image for classification of the object of interest (which contains the target object), and pair the frames together. We define a learning algorithm to find the feature mapping from the input frames to the target frames to improve the classification accuracy. Using the proposed algorithm, we obtain optimal classification accuracy, and use this improvement to optimize the MLVRP classification algorithm. Our evaluation shows the method performs better than other algorithms in all cases, including (1) using a loss function to estimate the learning rate for the classifier; (2) using a loss function to estimate the feature mapping of the object of interest (i.e. the weighted training set). Furthermore, we show these results can be used to improve the classification accuracy of our classification system, and thus show that this method can be used to automatically solve an MLVRP that involves a loss function.