SVM Algorithm

Artificial intelligence (AI) can use seismic training data to discover relationships between inputs and outcomes in real-world applications. Despite this, particularly when using geographical data, it has not been used to predict earthquakes in the Flores Sea. The algorithm will read the seismic data as a pattern of iterations throughout the operation. The output data is created by grouping based on clusters using the most effective WCSS analysis, while the input features are derived from the original international resource information system (IRIS) web service data. Given that earthquake prediction is an effort to reduce seismic disasters, this research is essential. By generating predictions, it can reduce the devastation caused by earthquakes. Using the support vector machine (SVM), hyperparameter support vector machine (HP-SVM), and particle swarm optimization support vector machine (PSO-SVM) algorithms, this study seeks to forecast the Flores Sea earthquake. According to the estimation results, the SVM algorithm's evaluation value is less precise than that of the HP-SVM, especially the linear HP-SVM and HP-SVM Polynomial models. However, the HP-SVM RBF model's accuracy rating is identical to that of the traditional SVM model. The improvement of the PSO-SVM model, which has the finest gamma position and a value of 9.

Cet article 1 propose une approche hybride pour la segmentation de documents basée sur l'agrégation de différentes solutions. Divers algorithmes de segmentation peuvent être utilisés dans le système, ce qui permet la combinaison de stratégies multiples (spécifiques au domaine, supervisées et nonsupervisées). Un ensemble de documents étiquetés, segmentés au préalable et représentatif du domaine ciblé, doit être fourni pour être utilisé comme ensemble d'entraînement pour l'apprentissage des méthodes supervisées, et aussi comme ensemble de test pour l'évaluation de la performance de chaque méthode, ce qui déterminera leur poids lors de la phase d'agrégation. L'approche proposée présente de bonnes performances dans un scénario expérimental issu d'un corpus extrait du domaine juridique.

Introduction: Leptodactylus latinasus and Physalaemus cuqui are sympatric anuran species with similar environmental requirements and contrasting reproductive modes. Climatic configuration determines distribution patterns and promotes sympatry of environmental niches, but specificity/selectivity determines the success of reproductive modes. Species distribution models (SDM) are a valuable tool to predict spatio-temporal distributions based on the extrapolation of environmental predictors. Objectives: To determine the spatio-temporal distribution of environmental niches and assess whether the protected areas of the World Database of Protected Areas (WDPA) allow the conservation of these species in the current scenario and future. Methods: We applied different algorithms to predict the distribution and spatio-temporal overlap of environmental niches of L. latinasus and P. cuqui within South America in the last glacial maximum (LGM), middle-Holocene, current and future scenarios. We assess the conservation status of both species with the WDPA conservation units. Results: All applied algorithms showed high performance for both species (TSS = 0.87, AUC = 0.95). The L. latinasus predictions showed wide environmental niches from LGM to the current scenario (49 % stable niches, 37 % gained niches, and 13 % lost niches), suggesting historical fidelity to stable climatic-environmental regions. In the current-future transition, L. latinasus would increase the number of stable (70 %) and lost (20 %) niches, suggesting fidelity to lowland regions and a possible trend toward microendemism. P. cuqui loses environmental niches from the LGM to the current scenario (25 %) and in the current-future transition (63 %), increasing the environmental sympathy between both species; 31 % spatial overlap in the current scenario and 70 % in the future.

Résumé. Nous proposons une nouvelle approche pour le calcul de similarité sémantique entre phrases en utilisant les noyaux sémantiques qui les composent. Ces noyaux, sous la forme de triplets (sujet, verbe et objet) sont supposés porteurs de l’information des phrases dont ils sont extraits. Sur la base de la comparaison sémantique de noyaux, on extrait un ensemble d’indicateurs descriptifs. Nous utilisons ensuite un apprentissage automatique, sur un benchmark contenant des phrases dont la similarité sémantique a été évaluée par des experts humains, afin de déterminer l’importance de chaque indicateur et de construire ainsi un modèle capable de fournir une mesure de similarité sémantique entre phrases. Les expérimentations et les études comparatives, effectuées avec d’autres approches permettant l’estimation des similarités sémantiques entre phrases, montrent les bonnes performances de notre approche. En se basant sur cette dernière, un outil de navigation sémantique est en cours de ...

In this article, we propose a homemade electronic nose to distinguish between two types of potatoes: the first type is traditionally treated with donkey and sheep manure, and the other type is treated with chicken manure. The proposed tool consists of a network of commercial metal oxide sensors, a data acquisition card, and a personal computer for data pre-processing and processing. Two methods were used, namely, support vector machines (SVM) and k-nearest neighbors (KNN) with 5-fold cross-validation and which achieved the same success rate of 97.5%. These results demonstrate that our concept, which is quick, simple, and inexpensive, can discriminate between potatoes based on the method of fertilization used in the field.

L’Extraction d’Information est une discipline qui a emerge du Traitement Automatique des Langues afin de proposer des analyses fines d’un texte ecrit en langage naturel et d’ameliorer la recherche d’informations specifiques. Les techniques d’extraction d’information ont enormement evolue durant les deux dernieres decennies.Les premiers systemes d’extraction d’information etaient des systemes a base de regles ecrites manuellement. L’ecriture manuelle des regles etant devenue une tâche fastidieuse, des algorithmes d’apprentissage automatique de regles ont ete developpes.Ces algorithmes necessitent cependant la redaction d’un guide d’annotation detaille, puis l’annotation manuelle d’une grande quantite d’exemples d’entrainement. Pour minimiser l’effort humain requis dans les deux familles d’approches de mise au point de regles, nous avons propose, dans ce travail de these, une approche hybride qui combine les deux en un seul systeme interactif qui procede en plusieurs iterations.Ce sys...

Crop prediction is the process of forecasting the yield or production of crops for a given period, based on historical data, weather patterns, and other relevant factors. The prediction can be used to inform decisions regarding planting, harvesting, and marketing of crops. Machine learning and artificial intelligence techniques are increasingly being used to improve crop prediction accuracy. These techniques use algorithms to analyze large amounts of data, such as weather patterns, soil conditions, and crop history, to make predictions about future crop yields. Crop prediction models can be used by farmers, agribusinesses, and governments to optimize crop management, reduce waste, and maximize profits. Accurate crop prediction can also help to mitigate the impact of climate change on agricultural production

Crop prediction is the process of forecasting the yield or production of crops for a given period, based on historical data, weather patterns, and other relevant factors. The prediction can be used to inform decisions regarding planting, harvesting, and marketing of crops. Machine learning and artificial intelligence techniques are increasingly being used to improve crop prediction accuracy. These techniques use algorithms to analyze large amounts of data, such as weather patterns, soil conditions, and crop history, to make predictions about future crop yields. Crop prediction models can be used by farmers, agribusinesses, and governments to optimize crop management, reduce waste, and maximize profits. Accurate crop prediction can also help to mitigate the impact of climate change on agricultural production

In this paper, we propose a comparative study between three approaches for automatic summarization of Arabic documents. Thus, we proposed three methods for extracting most representative sentences of a document. The first method is based on a symbolic ...

Crop prediction is the process of forecasting the yield or production of crops for a given period, based on historical data, weather patterns, and other relevant factors. The prediction can be used to inform decisions regarding planting, harvesting, and marketing of crops. Machine learning and artificial intelligence techniques are increasingly being used to improve crop prediction accuracy. These techniques use algorithms to analyze large amounts of data, such as weather patterns, soil conditions, and crop history, to make predictions about future crop yields. Crop prediction models can be used by farmers, agribusinesses, and governments to optimize crop management, reduce waste, and maximize profits. Accurate crop prediction can also help to mitigate the impact of climate change on agricultural production

La tâche de compréhension automatique du langage en contexte d'interaction (NLU pour Natural Language Understanding) est souvent réduite à la détection d'intentions et de concepts sur des corpus mono-domaines annotés avec une seule intention par énoncé. Afin de dépasser ce paradigme, nous cherchons à aborder des référentiels plus complexes en visant des représentations sémantiques structurées au-delà du simple modèle intention/concept. Nous nous intéressons au corpus MultiWOZ, couramment utilisé pour le suivi de l'état du dialogue. Nous questionnons la projection de ces annotations sémantiques complexes pour le NLU, en comparant plusieurs approches d'étiquetage de séquence, puis en proposant un nouveau formalisme inspiré des méthodes de génération de graphe pour la modélisation sémantique AMR. Nous discutons enfin le potentiel des approches génératives.

In this article, we propose a homemade electronic nose to distinguish between two types of potatoes: the first type is traditionally treated with donkey and sheep manure, and the other type is treated with chicken manure. The proposed tool consists of a network of commercial metal oxide sensors, a data acquisition card, and a personal computer for data pre-processing and processing. Two methods were used, namely, support vector machines (SVM) and k-nearest neighbors (KNN) with 5-fold cross-validation and which achieved the same success rate of 97.5%. These results demonstrate that our concept, which is quick, simple, and inexpensive, can discriminate between potatoes based on the method of fertilization used in the field.

Nous proposons dans cet article une approche d'extraction d'informations, basée sur une ontologie, et appliquée à des documents issus de catalogues publicitaires. Les documents des catalogues sont des descriptifs de produits relativement pauvres. Les informations à extraire, ou annotations, concernent les catégories et les caractéristiques des produits considérés, répertoriées dans une ontologie de domaine. L'extraction d'informations concernant un produit consiste en un peuplement de cette ontologie, plus précisément le peuplement des concepts représentant ses catégories et ses caractéristiques. La relative pauvreté des descriptifs rend irréalisable un peuplement totalement automatique. Nous proposons donc une approche en deux étapes : (1) une première étape d'annotation semi-automatique qui porte sur un petit ensemble de documents ; (2) une deuxième étape qui annote l'ensemble des autres documents de façon entièrement automatique, en s'appuyant sur des mécanismes d'apprentissage automatique exploitant les résultats de la première étape. L'originalité de ce travail consiste en une approche incrémentale de raffinement des informations extraites. Le travail décrit a été appliqué sur des jeux de données réelles concernant des jouets. ABSTRACT. In this paper, we propose an approach of information extraction, based on an ontology, and applied to documents from advertising catalogs. Documents are relatively poor descriptions of products. The information to be extracted, or annotations, concern the categories and features of the products, listed in a domain ontology. Thus, the information extraction about a product is actually an ontology population process, more precisely the population of concepts representing its categories and features. The poverty of the descriptions makes a fully automatic population impossible. We propose a two-step approach: (1) a first semi-automatic annotation step, which covers a small set of documents; (2) a second step, which annotates all other documents, in an entirely automatic way, based on machine learning mechanisms exploiting the results of the first step. The originality of this work relies on an incremental approach Revue d'intelligence artificielle-n o 5/2016, 557-578 MOTS-CLÉS : extraction d'informations, peuplement d'ontologie, annotation sémantique, application dans le domaine du e-commerce.

Inherited retinal diseases cause severe visual deficits in children. They are classified in outer and inner retina diseases, and often cause blindness in childhood. The diagnosis for this type of illness is challenging, given the wide range of clinical and genetic causes (with over 200 causative genes). It is routinely based on a complex pattern of clinical tests, including invasive ones, not always appropriate for infants or young children. A different approach is thus needed, that exploits Chromatic Pupillometry, a technique increasingly used to assess outer and inner retina functions. This paper presents a novel Clinical Decision Support System (CDSS), based on Machine Learning using Chromatic Pupillometry in order to support diagnosis of Inherited retinal diseases in pediatric subjects. An approach that combines hardware and software is proposed: a dedicated medical equipment (pupillometer) is used with a purposely designed custom machine learning decision support system. Two distinct Support Vector Machines (SVMs), one for each eye, classify the features extracted from the pupillometric data. The designed CDSS has been used for diagnosis of Retinitis Pigmentosa in pediatric subjects. The results, obtained by combining the two SVMs in an ensemble model, show satisfactory performance of the system, that achieved 0.846 accuracy, 0.937 sensitivity and 0.786 specificity. This is the first study that applies machine learning to pupillometric data in order to diagnose a genetic disease in pediatric age

Mes dédicaces ne sont que l'expression de mes profondes gratitudes, de mes salutations chaleureuses et de ma sincère reconnaissance à tous ceux qui comblent ma vie et y confèrent son goût et sa saveur. Je dédie cette thèse à : Mon père Ce symbole de sacrifice et de dévouement, ses aides et ses recommandations m'ont souvent incité à persévérer dans l'effort et à progresser dans ma vie universitaire et professionnelle. Ma mère Ce rayon de soleil qui ne cesse d'éclairer ma vie, cette source inépuisable d'amour, de tendresse et d'affection, grâce à ses conseils, j'ai pu frayer mon chemin et savourer le goût de la réussite et du succès.

Iskandar Keskes Mohamed Mahdi Boudabous Mohamed Hédi Maaloul Lamia Hadrich Belguith (1) ANLP Research Group, Laboratoire MIRACL, Route de Tunis Km 10, BP 242, Sfax, Tunisie (2) Laboratoire IRIT, 118 Route de Narbonne, F-31062 Toulouse Cedex 9, France (3) Laboratoire LPL, 5 avenue Pasteur, BP 80975, 13604 Aix-en-Provence, France Keskes@irit.fr, mehdiboudabous@gmail.com mohamed.maaloul@lpl-aix.fr, l.belguith@fsegs.rnu.tn

The purpose of this paper is to predict stock prices using Hybrid GA-SVM Algorithm. Predicting time series such as stock price forecasting is one of the most important issues in financial field. In real life, identifying time series movements in stock price indices is very complex. Therefore, the use of a classical model alone cannot accurately predict stock price indices. Hence, by using combined methods, uncertainty in forecasting can be reduced. In stock price forecasting in financial sector, more than 100 indicators have been created to understand stock market behavior, so, identifying the appropriate indicators is a challenging problem. One of the techniques that has recently been studied for serial forecasting is support regression Vector (SVR) or machine support vector (SVM). This study uses the GA-SVM hybrid algorithm to predict the stock price index. Experimental results show that Hybrid GA-SVM Algorithm provides a more appropriate and promising alternative to stock market forecasting.