This paper describes our proposed solutions designed for a STS core track within the SemEval 2016 English Semantic Textual Similarity (STS) task. Our method of similarity detection combines recursive autoencoders with a WordNet award-penalty system that accounts for semantic relatedness, and an SVM classifier, which produces the final score from similarity matrices. This solution is further sup...

The importance of metrics in machine learning has attracted a growing interest for distance and similarity learning, and especially the Mahalanobis distance. However, it is worth noting that this research field lacks theoretical guarantees that can be expected on the generalization capacity of the classifier associated to a learned metric. The theoretical framework of ( , γ, τ)-good similarity ...

Most of the existing Non-Cooperative Target Recognition (NCTR) systems follow “closed world” assumption, i.e., they only work with what was previously observed. Nevertheless, real world is relatively “open” in sense that knowledge environment incomplete. Therefore, unknown targets can feed recognition system at any time while it operational. Addressing this issue, Openmax classifier has been re...

We propose a winner-takes-all (WTA) classifier for structures represented by graphs. WTA classification follows the principle elimination of competition. The input structure is assigned to the class corresponding to the winner of the competition. In experiments we investigate the performance of the WTA classifier and compare it with the canonical maximum similarity (MS) classifier.

t his paper presents a new feature selection approach for automatically extracting   ms lesions in 3d mr images. presented method is applicable to different types of ms lesions. in this method, t1, t2 and flair images are firstly preprocessed. in the next phase, effective features to extract ms lesions are selected by using a genetic algorithm. the fitness function of the genetic algorithm is t...

In this paper we propose a wet lab algorithm for prediction of radiation fog by DNA computing. The concept of DNA computing is essentially exploited for generating the classifier algorithm in the wet lab. The classifier is based on a new concept of similarity based fuzzy reasoning suitable for wet lab implementation. This new concept of similarity based fuzzy reasoning is different from convent...

This paper addresses the problem of Near Duplicate document. Propose a new method to detect near duplicate document from a large collection of document set. This method is classified into three steps. Feature selection, similarity measures and discriminant function. Feature selection performs pre-processing; calculate the weight of each terms and heavily weighted term is selected as a features ...

The importance of metrics in machine learning has attracted a growing interest for distance and similarity learning. We study here this problem in the situation where few labeled data (and potentially few unlabeled data as well) is available, a situation that arises in several practical contexts. We also provide a complete theoretical analysis of the proposed approach. It is indeed worth noting...