Search results for "Perceptron"
showing 9 items of 89 documents
Fast and Robust Face Detection on a Parallel Optimized Architecture implemented on FPGA
2009
In this paper, we present a parallel architecture for fast and robust face detection implemented on FPGA hardware. We propose the first implementation that meets both real-time requirements in an embedded context and face detection robustness within complex backgrounds. The chosen face detection method is the Convolutional Face Finder (CFF) algorithm, which consists of a pipeline of convolution and subsampling operations, followed by a multilayer perceptron. We present the design methodology of our face detection processor element (PE). This methodology was followed in order to optimize our implementation in terms of memory usage and parallelization efficiency. We then built a parallel arch…
influence of raw data analysis for the use of neural networks for wind farm productivity prediction
2011
In the last decade wind energy had a strong growth because of cost effectiveness of the technology and the high remunerative of investments. The increase of wind power penetration in power grids, however, makes necessary the development of instruments for prediction of productivity of a wind farm. This paper presents a study dealing with the capability of neural network to forecast short term production of a wind farm by the correlation of wind and energy production data. Available measures of wind parameters were related to productivity data of a real wind farm. Also wind data not strictly related to the site have been used in order to assess their possible influence on the production. Aft…
Multiple Classifiers and Data Fusion for Robust Diagnosis of Gearbox Mixed Faults
2019
Detection and isolation of single and mixed faults in a gearbox are very important to enhance the system reliability, lifetime, and service availability. This paper proposes a hybrid learning algorithm, consisting of multilayer perceptron (MLP)- and convolutional neural network (CNN)-based classifiers, for diagnosis of gearbox mixed faults. Domain knowledge features are required to train the MLP classifier, while the CNN classifier can learn features itself, allowing to reduce the required knowledge features for the counterpart. Vibration data from an experimental setup with gearbox mixed faults is used to validate the effectiveness of the algorithms and compare them with conventional metho…
Dziļo neironu tīklu veidi un galvenās risināmās problēmas
2017
Darbā tika pētīta mašīnmācīšanās apakšnozare, dziļie neironu tīkli. Lai gūtu nepieciešamās pamatzināšanas par neironu tīklu uzbūvi un darbības principiem, sākotnēji tika apskatīti vienkāršākie neironu tīkli – perceptroni. Pēc tam tika pētīta dziļā mācīšanās un dziļie neironu tīkli, to, kas tiek saprasts ar šo jēdzienu, kādā veidā tie atšķiras no tradicionālajiem risinājumiem, kādi ir to veidi un to kādus uzdevumus un kādā veidā tie spēj risināt. Rezultātā autors ieguva zināšanas neironu tīklu nozarē, kuras var izmantot tālākajā izpētē. Ir sniegts ieskats neironu tīklu pamatos, izpētīta dziļo neironu tīklu nozare, apskatīti populārākie dziļie neironu tīkli un to dažādais pielietojums. Kā arī…
Prediction of active peak force using a multilayer perceptron
2017
Both kinematic parameters and ground reaction forces (GRFs) are necessary for understanding the biomechanics of running. Kinematic information of a runner is typically measured by a motion capture system whereas GRF during the support phase of running is measured by force platforms. To analyze both kinematics and kinetics of a runner over several subsequent contacts, an instrumented treadmill or alternatively several force platforms installed over a regulated space are available options, but they are highly immovable, expensive, and sometimes even impractical options. Naturally, it would be highly useful to predict GRFs using a motion capture system only and this way reduce costs and comple…
Multilayer perceptron training with multiobjective memetic optimization
2016
Machine learning tasks usually come with several mutually conflicting objectives. One example is the simplicity of the learning device contrasted with the accuracy of its performance after learning. Another common example is the trade-off that must often be made between the rate of false positive and false negative predictions in diagnostic applications. For computer programs that learn from data, these objectives are formulated as mathematical functions, each of which describes one facet of the desired learning outcome. Even functions that intend to optimize the same facet may behave in a subtly different and mutually conflicting way, depending on the task and the dataset being examined. Mul…
Neural Networks Ensemble for Cyclosporine Concentration Monitoring
2001
This paper proposes the use of neural networks ensemble for predicting the cyclosporine A (CyA)concen tration in kidney transplant patients. In order to optimize clinical outcomes and to reduce the cost associated with patient care, accurate prediction of CyA concentrations is the main objective of therapeutic drug monitoring. Thirty-two renal allograft patients and different factors (age, weight, gender, creatinine and post-transplantation days, together with past dosages and concentrations)w ere studied to obtain the best models. Three kinds of networks (multilayer perceptron, FIR network, Elman recurrent network) and the formation of neural-network ensembles were used. The FIR network, y…
Mašīntulkotu nosaukto entitāšu gramatisko locījumu noteikšana automātiskajā pēcrediģēšanā
2020
Dziļās mašīnmācīšanās sasniegumi ir veicinājuši ievērojamu attīstību mašīntulkošanas jomā. Tomēr joprojām bieži tiek novērots, ka retas nosauktās entitātes neironu mašīntulkošanas sistēmas iztulko nepareizi. Kļūdaini tulkotu nosaukto entitāšu labošanu var veikt, izmantojot automātisko pēcrediģēšanu. Pēcrediģēšana kā vienu no soļiem ietver teikuma kontekstā iederīgas morfoloģiskās formas noteikšanu. Šī darba mērķis bija izstrādāt un apmācīt klasifikatoru nosauktās entitātes teikumā iederīgo morfoloģisko kategoriju formu paredzēšanai. Darba ietvaros tika izpētīts, kā risinātas līdzīgas ar vārdu kontekstuālu morfoloģisko analīzi saistītas problēmas, sagatavoti klasifikatoram piemēroti apmācība…
Evaluating the performance of artificial neural networks for the classification of freshwater benthic macroinvertebrates
2014
Abstract Macroinvertebrates form an important functional component of aquatic ecosystems. Their ability to indicate various types of anthropogenic stressors is widely recognized which has made them an integral component of freshwater biomonitoring. The use of macroinvertebrates in biomonitoring is dependent on manual taxa identification which is currently a time-consuming and cost-intensive process conducted by highly trained taxonomical experts. Automated taxa identification of macroinvertebrates is a relatively recent research development. Previous studies have displayed great potential for solutions to this demanding data mining application. In this research we have a collection of 1350 …