The study of microplastic pollution involves multidisciplinary analyses on a large number of microplastics. Therefore, providing an overview of plastic pollution is time consuming and, despite high throughput analyses, remains a major challenge. The objective of this study is to propose a protocol to determine how many microplastics must be analyzed to give a representative view of the particle size distribution and chemical nature, and calculate the associated margin error. Based on microplastic data from Tara Mediterranean campaign, this approach is explained through different examples. In this particular case, the results show that only 3% of the collected microplastics need to be analyzed to give a precise view on the scale of the North West Mediterranean Basin (error <5%), and 17.7% to give an overview manta per manta (error <10%). This approach could be an important practical contribution to microplastic studies.

No abstract

5-Hydroxytryptamine subtype-4 (5-HT 4 ) receptors have stimulated considerable interest amongst scientists and clinicians owing to their importance in neurophysiology and potential as therapeutic targets. A comparative analysis of hierarchical methods applied to data from one thousand 5-HT 4 receptor -ligand binding interactions was carried out. The chemical structures were described as chemical and pharmacophore fingerprints. The definitions of indices, related to the quality of the hierarchies in being able to distinguish between active and inactive compounds, revealed two interesting hierarchies with the Unity (1 active cluster) and pharmacophore fingerprints (4 active clusters). The results of this study also showed the importance of correct choice of metrics as well as the effectiveness of a new alternative of the Ward clustering algorithm named Energy (Minimum E-Distance method). In parallel, the relationship between these classifications and a previously defined 3D 5-HT 4 antagonist pharmacophore was established.

The potential of quantile regression (QR) and quantile support vector machine regression (QSVMR) was analyzed for the definitions of quantitative structure-activity relationship (QSAR) models associated with a diverse set of chemicals toward a particular endpoint. This study focused on a specific sensitive endpoint (acute toxicity to algae) for which even a narcosis QSAR model is not actually clear. An initial dataset including more than 401 ecotoxicological data for one species of algae (Selenastrum capricornutum) was defined. This set corresponds to a large sample of chemicals ranging from classical organic chemicals to pesticides. From this original data set, the selection of the different subsets was made in terms of the notion of toxic ratio (TR), a parameter based on the ratio between predicted and experimental values. The robustness of QR and QSVMR to outliers was clearly observed, thus demonstrating that this approach represents a major interest for QSAR associated with a diverse set of chemicals. We focused particularly on descriptors related to molecular surface properties.

The development of connected devices and their daily use are today at the origin of the omnipresence of Wi-Fi wireless networks. However, these Wi-Fi networks are often vulnerable, and can be used by malicious people to disturb services, intercept sensitive data or to gain access to system. In railways, trains are now equipped with wireless communication systems for operational purposes or for passenger services. In both cases, defense strategies have to be developed to prevent misuses of the networks. The first objective of this study is to propose a monitoring solution, which is independent of the communication networks, to detect the occurrence of attacks. The second objective is to develop a method able to classify attacks of different types: the intentional electromagnetic interference (IEMI), i.e., jamming attacks, and the protocol-based attacks. This study focuses on the IEEE 802.11n Wi-Fi protocol. To perform these analyses, we propose to monitor and to analyze electromagnetic (EM) signals received by a monitoring antenna and a receiver collecting EM spectra. After that, we build a classification protocol following two steps: the first consists in the construction of a Support Vector Machine (SVM) classification model using the collected spectra and the second step uses this SVM model to predict the class of the attack (if any). A timebased correction of this prediction using the nearest neighbors is also included in this second step.

No abstract

No abstract

Extreme Learning Machine (ELM) technology has started gaining interest in the channel estimation and equalization aspects of wireless communications systems. This is due to its fast training and global optimization capabilities that might allow the ELM-based receivers to be deployed in an online mode while facing the channel scenario at hand. However, ELM still needs a relatively large amount of training samples, thus causing important losses in spectral resources. In this work, we make use of the ensemble learning theory to propose different ensemble learning-based ELM equalizers that need much less spectral resources, while achieving better performance accuracy. Also, we verify the robustness of our proposed equalizers within different communication settings and channel scenarios by conducting different Monte Carlo simulations.