Search results for "Neural"
showing 10 items of 2783 documents
Artificial neural network applied to the discrimination of antibacterial activity by topological methods
2000
Abstract A new topological method that makes it possible to discriminate the active and inactive molecules on the basis of their chemical structures is applied in the present study to the antibacterial agents. This method uses neural networks in which training algorithms are used as well as different concepts and methods of artificial intelligence with a suitable set of topological descriptors. It is possible to obtain a QSAR interpretation of the information contained in the network after the training has been carried out.
Application of adaptive models for the determination of the thermal behaviour of a photovoltaic panel
2013
The use of reliable forecasting models for the PV temperature is necessary for a more correct evaluation of energy and economic performances. Climatic conditions certainly have a remarkable influence on thermo-electric behaviour of the PV panel but the physical system is too complex for an analytical representation. A neural-network-based approach for solar panel temperature modelling is here presented. The models were trained using a set of data collected from a test facility. Simulation results of the trained neural networks are presented and compared with those obtained with an empirical correlation.
Extraction of Endmembers from Spectral Mixtures
1999
Abstract Linear spectral mixture modeling (LSMM) divides each ground resolution element into its constituent materials using endmembers which represent the spectral characteristics of the cover types. However, it is difficult to identify and estimate the spectral signature of pure components or endmembers which form the scene, since they vary with the scale and purpose of the study. We propose three different methods to estimate the spectra of pure components from a set of unknown mixture spectra. Two of the methods consist in different optimization procedures based on objective functions defined from the coordinate axes of the dominant factors. The third one consists in the design of a neu…
Reti Neurali per la Realizzazione di Mappe di Suscettibilità per il Rischio Frana
2012
La valutazione della suscettibilità al dissesto rimane uno degli approcci più utilizzanti e più efficaci per l'analisi della pericolosità da frana. Come noto, la correlazione tra il fenomeno fisico ed i fattori predisponenti sulla base degli eventi accaduti in passato è il punto chiave di tale analisi. I metodi statistici, uniti con le tecnologie GIS, si sono rilevati in questi anni tra gli strumenti più idonei e più efficaci per la valutazione e la modellazione di tale correlazione. Tuttavia, questi metodi richiedono spesso ipotesi restrittive circa la distribuzione statistica dei dati che spesso non vengono rispettate. Per tale motivo si sono anche sviluppate metodologie alternative basat…
Sistemi di monitoraggio meteorologico per l'analisi del campo di precipitazione in aree urbane finalizzati al preannuncio precoce di dissesti idrogeo…
2018
Lo sviluppo dei sistemi informativi territoriali, insieme con quello di avanzate tecniche di monitoraggio ambientale, potrebbe risultare molto utile nell'ambito del preavviso e/o della mitigazione del rischio idrogeologico. Questo lavoro presenta un sistema integrato per il preannuncio del rischio idrogeologico il cui punto di forza risiede in un sistema di monitoraggio climatico e idrologico/geotecnico costituito da diversi sensori wireless che registrano una serie di informazioni relative a grandezze di tipo meteorologico, idrologico e geotecnico. I dati registrati e inviati ad una piattaforma web sono utilizzati per alimentare una rete neurale artificiale, opportunamente creata e addestr…
Predicting Real-Time Roadside CO and NO2 Concentrations using Neural Networks
2008
Prediction of concrete materials compressive strength using surrogate models
2022
Using soft computing methods could be of great interest in predicting the compressive strength of Ultra-High-Performance Fibre Reinforced Concrete (UHPFRC). Therefore, this study developed four soft computing techniques. The models are the Linear- relationship (LR), pure quadratic, M5P-tree (M5P), and artificial neural network (ANN). The models were trained and developed using 306 datasets comprising 11 input parameters, including the curing temperature (T), the water-to-cement ratio (w/c), silica fume (SF), cement content (C), fiber content (Fb), water (W), sand content (S), superplasticizer (SP), fiber aspect ratio (AR) and curing time (t). Experimental results were used and compared to t…
SURFACE ROUGHNESS PREDICTION OF ELECTRO-DISCHARGE MACHINED COMPONENTS USING ARTIFICIAL NEURAL NETWORKS
2016
Electro-Discharge machining (EDM) is a thermal process comprising a complex metal removal mechanism, which involves the formation of a plasma channel between the tool and the workpiece electrodes leading to the melting and evaporation of the material to be removed. EDM is considered especially suitable for machining complex contours with high accuracy, as well as for materials that are not amenable to conventional removal methods. However, several phenomena negatively affecting the surface integrity of EDMed workpieces need to be taken into account and studied in order to achieve the optimization of the process. Recently, artificial neural networks (ANN) have emerged as a novel modeling tec…
Revealing the nature of metakaolin-based concrete materials using artificial intelligence techniques
2022
In this study, a model for the estimation of the compressive strength of concretes incorporating metakaolin is developed and parametrically evaluated, using soft computing techniques. Metakaolin is a component extensively employed in recent decades as a means to reduce the requirement for cement in concrete. For the proposed models, six parameters are accounted for as input data. These are the age at testing, the metakaolin percentage in relation to the total binder, the water-to-binder ratio, the percentage of superplasticizer, the binder to sand ratio and the coarse to fine aggregate ratio. For training and verification of the developed models a database of 867 experimental specimens has …
An intelligent way to predict the building thermal needs: ANNs and optimization
2022
The evaluation of the energy performance of existing or new buildings is a fundamental action to guarantee the feasibility of a project and the achievement of the minimum efficiency requirements. In general, the determination of the thermal loads of a building is carried out via software but their use requires adequate knowledge of physical phenomena and therefore the presence of an expert user. Furthermore, the resolution can be difficult to implement and can require high computational costs; all conditions that can influence the success of a project. Based on these considerations, this work proposes an alternative solution to traditional calculation tools, which in a simple and effective …