Downloads & Free Reading Options - Results

This paper presented neural network based maximum power point tracking on the design of photovoltaic power input to a DC DC boot converter to the load. Simulink model of photovoltaic array tested the neural network with different temperature and irradiance for maximum power point of a photovoltaic system. DC DC boot converter is used in load when an average output voltage is stable required which can be lower than the input voltage. At the end, the different temperature and irradiance of the data collected from the photovoltaic array system is used to train the neutral network and output efficiency of the designed DC DC boot converter with MPPT control strategy is accepted the maximum power amount to show the result voltage, current and power output for each different have been presented. And also demonstrated that the neural network based MPPT tracking require less time and more accurate results than the other algorithm based MPPT. By Myint Thuzar | Cho Hnin Moh Moh Aung "Artificial Neural Network for Solar Photovoltaic System Modeling and Simulation" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-3 | Issue-5 , August 2019, URL: https://www.ijtsrd.com/papers/ijtsrd27867.pdf Paper URL https://www.ijtsrd.com/engineering/electrical-engineering/27867/artificial-neural-network-for-solar-photovoltaic-system-modeling-and-simulation/myint-thuzar

Since the 1970s, most airlines have incorporated computerized support for managing disruptions during flight schedule execution. However, existing platforms for airline disruption management (ADM) employ monolithic system design methods that rely on the creation of specific rules and requirements through explicit optimization routines, before a system that meets the specifications is designed. Thus, current platforms for ADM are unable to readily accommodate additional system complexities resulting from the introduction of new capabilities, such as the introduction of unmanned aerial systems (UAS), operations and infrastructure, to the system. To this end, we use historical data on airline scheduling and operations recovery to develop a system of artificial neural networks (ANNs), which describe a predictive transfer function model (PTFM) for promptly estimating the recovery impact of disruption resolutions at separate phases of flight schedule execution during ADM. Furthermore, we provide a modular approach for assessing and executing the PTFM by employing a parallel ensemble method to develop generative routines that amalgamate the system of ANNs. Our modular approach ensures that current industry standards for tardiness in flight schedule execution during ADM are satisfied, while accurately estimating appropriate time-based performance metrics for the separate phases of flight schedule execution.

Natural Language Processing (NLP) systems commonly leverage bag-of-words co-occurrence techniques to capture semantic and syntactic word relationships. The resulting word-level distributed representations often ignore morphological information, though character-level embeddings have proven valuable to NLP tasks. We propose a new neural language model incorporating both word order and character order in its embedding. The model produces several vector spaces with meaningful substructure, as evidenced by its performance of 85.8% on a recent word-analogy task, exceeding best published syntactic word-analogy scores by a 58% error margin. Furthermore, the model includes several parallel training methods, most notably allowing a skip-gram network with 160 billion parameters to be trained overnight on 3 multi-core CPUs, 14x larger than the previous largest neural network.

In this paper, we propose convolutional neural networks for learning an optimal representation of question and answer sentences. Their main aspect is the use of relational information given by the matches between words from the two members of the pair. The matches are encoded as embeddings with additional parameters (dimensions), which are tuned by the network. These allows for better capturing interactions between questions and answers, resulting in a significant boost in accuracy. We test our models on two widely used answer sentence selection benchmarks. The results clearly show the effectiveness of our relational information, which allows our relatively simple network to approach the state of the art.

In this paper, we propose convolutional neural networks for learning an optimal representation of question and answer sentences. Their main aspect is the use of relational information given by the matches between words from the two members of the pair. The matches are encoded as embeddings with additional parameters (dimensions), which are tuned by the network. These allows for better capturing interactions between questions and answers, resulting in a significant boost in accuracy. We test our models on two widely used answer sentence selection benchmarks. The results clearly show the effectiveness of our relational information, which allows our relatively simple network to approach the state of the art.

Neural networks provide an alternative method of building models of human performance. They can learn behavior from examples, reducing the need for many identical repetitions and intensive analysis. A properly trained net can be very robust in its response to a novel stimulus. This opens the door to modeling performance in the presence of an interactive stimulus. Neural networks provide the possibility of robust models that can operate interactively in real time, depending on the size and architecture of the net and the application. A neural network architecture derived from recurrent back propagation is presented which learn to mimic human behavior and performance in a sample task. It shows operating characteristics similar to those of human subjects, and even makes the same kinds of mistakes. Possible application are discussed. (js)

Neural networks provide an alternative method of building models of human performance. They can learn behavior from examples, reducing the need for many identical repetitions and intensive analysis. A properly trained net can be very robust in its response to a novel stimulus. This opens the door to modeling performance in the presence of an interactive stimulus. Neural networks provide the possibility of robust models that can operate interactively in real time, depending on the size and architecture of the net and the application. A neural network architecture derived from recurrent back propagation is presented which learn to mimic human behavior and performance in a sample task. It shows operating characteristics similar to those of human subjects, and even makes the same kinds of mistakes. Possible application are discussed. (js)

In the work the approach to neural networks application is discussed as the competitive computing technology in the analytical information processing. The general function chart and algorithm of transformation of the analytical information are developed within the framework of a competition principle. The basic attention is given to a choice of topology of a neural network both its structural and parametrical synthesis. 

Trajectory reconstruction of pedestrian is of paramount importance to understand crowd dynamics and human movement pattern, which will provide insights to improve building design, facility management and route planning. Camera-based tracking methods have been widely explored with the rapid development of deep learning techniques. When moving to indoor environment, many challenges occur, including occlusions, complex environments and limited camera placement and coverage. Therefore, we propose a novel indoor trajectory reconstruction method using building information modeling (BIM) and graph neural network (GNN). A spatial graph representation is proposed for indoor environment to capture the spatial relationships of indoor areas and monitoring points. Closed circuit television (CCTV) system is integrated with BIM model through camera registration. Pedestrian simulation is conducted based on the BIM model to simulate the pedestrian movement in the considered indoor environment. The simulation results are embedded into the spatial graph for training of GNN. The indoor trajectory reconstruction is implemented as GNN conducts edge classification on the spatial graph

Trajectory reconstruction of pedestrian is of paramount importance to understand crowd dynamics and human movement pattern, which will provide insights to improve building design, facility management and route planning. Camera-based tracking methods have been widely explored with the rapid development of deep learning techniques. When moving to indoor environment, many challenges occur, including occlusions, complex environments and limited camera placement and coverage. Therefore, we propose a novel indoor trajectory reconstruction method using building information modeling (BIM) and graph neural network (GNN). A spatial graph representation is proposed for indoor environment to capture the spatial relationships of indoor areas and monitoring points. Closed circuit television (CCTV) system is integrated with BIM model through camera registration. Pedestrian simulation is conducted based on the BIM model to simulate the pedestrian movement in the considered indoor environment. The simulation results are embedded into the spatial graph for training of GNN. The indoor trajectory reconstruction is implemented as GNN conducts edge classification on the spatial graph

Ionospheric models have been developed to interpret Relocatable Over-the-Horizon Radar data. This thesis examines the applicability of neural networks to ionospheric modeling in support of Relocatable Over-the-Horizon Radar. Two neural networks were used for this investigation. The flrst network was trained and tested on experimental ionospheric sounding data. Results showed neural networks are excellent at modeling ionospheric data for a given day. The second network was trained on ionospheric models and tested on experimental data. Results showed neural networks are able to learn many ionospheric models and the modeling network generally agreed with the experimental data.

Introduction Presently, the loss of ground water levels and the increase in dissolved salts have given importance to the determination of salinity and the management of their variations in irrigated farms. Soil electrical conductivity is an indirect method to measure soil salts. The direct electrode contact method (Wenner method) is one of the widely used methods to rapidly measure soil ECa in farms. However, soil scientists prefer soil actual electrical conductivity (saturated extract electrical conductivity) (ECe) as an indicator of soil salinity, though its measurement is only possible in the laboratory. The aim of this study was to find a relationship between the prediction of soil actual electrical conductivity (ECe) in terms of temperature, moisture, bulk density and apparent electrical conductivity of soil (ECa). Thereby, the estimation of ECe would allow the partial calculation of ECa that is dependent upon soil salinity and dissolved salts. Materials and Methods This study used RBF neural network in Box-Behnken statistical design to explore the impacts of effective parameters on direct contact method in the measurement of soil ECa and provided a model to estimate ECe from ECa, temperature, moisture content and bulk density. In this study soil apparent electrical conductivity (ECa) was measured by direct contact (Wenner) method. The present study considered four most effective factors: ECa (saturated paste extract EC), moisture, bulk density, and temperature (Baradaran Motie et al., 2010). Given the characteristics of farming soils in Khorasan Razavi Province (Iran), the maximum and minimum of each independent variable were assumed as 0.5-6 mS.cm-1 for ECe, 5-25% for moisture content, 1-1.8 g.cm -3 for bulk density, and 2-37°C for soil temperature. Considering the experimental design, three moisture levels (5, 15 and 25%), three salinity levels (0.5, 3.25 and 6 mS.cm -1 ), three temperature levels (2, 19 and 37°C) and three compaction levels with bulk densities of 1, 1.4 and 1.8 g.cm -3 were assumed in 27 trials with predetermined arrangement on the basis of Box-Behnken technique. 13 common algorithms were explored in MATLAB software package for the training of the artificial neural network in order to find the optimum algorithm (Table 4). The input layer of the network designed by integrating a Randomized Complete Block Design (RCBD) with k-fold cross-validation. Using k-fold cross-validation, 20 different datasets were generated for training and validation of RBF neural network. Results and Discussion A combination of an RCBD and k-fold cross-validation was used. The results of both training and validation phases should be considered in the selection of training algorithm. In addition, R 2 of T1 training algorithm had a much lower standard deviation than other training algorithms. The lower standard deviation is, the more capable the algorithm would be in learning from different datasets. Considering all aspects, trainbr (T2) training algorithm was found to have the best performance among all 13 training algorithms of the neural network. Table 7 tabulates the results of means comparison for R2 of RBF model for both training and validation phases resulted from the application of some combinations of S and L2 factors as interaction. As can be observed, R 2 = 0.99 for all of them with no significant difference. However, the magnitude of order differed between training and validation phases. Given the importance of the training phase, L2=9 and S=0.1 were regarded as the optimum values. The sensitivity analysis of the network revealed that soil ECa, moisture, bulk density, and temperature had the highest to lowest impact on the estimation of soil ECe, respectively. This model can improve the precision of soil ECa measurement systems in the estimation and preparation of soil salinity maps. Furthermore, this model can save in time of data analysing and soil EC mapping because it does not need data recollection for the calibration of systems. A validation prose was done with a 60 field collected data set. The results of validation show R 2 =0.986 between predicted and measured ECa. Conclusion The present research focused on improving the precision of soil ECe measurement on the basis of easily accessible parameters (ECa, temperature, moisture, and bulk density). In conventional methods of soil EC mapping, the systems only measure soil ECa and then calibrate it to ECe by collecting some samples and using statistical methods. In this study, Soil ECe was estimated with R 2 = 0.99 by a multivariate artificial neural network model with the inputs, including ECa, temperature, moisture, and bulk density of soil without any need to collect further soil samples and calibration process. The Bayesian training algorithm was introduced as the best training algorithm for this neural network. Thereby, soil EC variation maps can be prepared with higher precision to estimate the spatial spread of salinity in farms. Also, the results imply that soil ECa, moisture, bulk density and temperature have the highest to lowest effectiveness on the estimation of soil ECe, respectively.

Introduction Tillage is a very important operation that influences the growth and productivity of agricultural products. It is necessary to introduce some conditions to improve soil physical properties, aeration, permeability and root development in tillage operations. However, in primary tillage, especially when moldboard ploughs are used, this may be time consuming and costly for researchers to use it in their research. Some researchers use physical experiments to perform the work, which the accuracy of the results is dependent on the measuring instruments precision. However, some other researchers use simulation and mathematical modeling to reduce the time and costs and increase the relative accuracy of the research results. Many studies have also shown that modeling the forces involved in tillage is a good way to estimate the performance of different tillage tools and improve their geometry. However, the key to success in numerical simulation of tillage operations is to simulate the exact instrumentation, based on the correct assumptions as well as the proper methods. The prediction of the forces involved in tillage tools has an important role in their design. Collecting data on the forces involved in tillage tool under different farm conditions is a time consuming and costly task. Therefore, the prediction of a tillage tool forces is very important for the designer and the user in order to achieve better performance of the tool. Materials and Methods In this study, a cylindrical moldboard made by Alpler Company in Turkey was used to simulate the moldboard. A measuring device was designed and constructed to measure the various points of the desired moldboard. Then, the spatial points obtained by the measuring device were presented to the SolidWorks 2016 software and the desired moldboard was modeled. The finite element method by Abacus 2016 was then used to simulate the interaction between soil and moldboard. Treatments used in simulated tillage operations included tillage depths (5, 10, 15, 20 and 25 cm) and forward speed (1, 1.5, 2, 2.5 and 3 millimeters per second). The independent variables were considered as tensile, vertical and lateral forces (Kilo newton). After simulating the tillage operations, tensile, vertical and lateral forces were obtained. These forces were modeled using response surface and artificial neural networks techniques. Then, the obtained models were compared using R2, RMSE and MRDM statistical indices and the best model was selected. Results and Discussion When using the response surface method, the quadratic model was selected by using the maximum value of the statistical indices R2, R2a and R2p, among the linear, two-factor and quadratic models. Then, the significance of model variables was evaluated by using variance analysis. The forces were also modeled by using the neural network method. According to the fitting curves and statistical indices of R2, RMSE and MRDM for the tensile, vertical and lateral forces, it is revealed that both methods could well predict the forces but artificial neural network was more suitable than the response surface method. Moreover, by investigating the interactions of tillage treatments and forward speed on the forces in this research, it was observed that by increasing the depth of tillage and velocity, tensile, vertical and lateral forces were increased nonlinearly by 66.55%, 68.47%, and 64.76%, respectively. Conclusion Regarding all the results obtained from this study, it can be concluded that the developed models using the artificial neural network in this research was a good and powerful tool for predicting the forces involved in moldboard ploughs both in the field operations and in related studies. It is also recommended that the developed models in this study can be used to manage the tillage operations, such as selecting the proper tractor. However, it is also suggested that other affecting factors, such as moldboard angles, should be included in future models to increase the ability of the model to predict the forces involved in moldboard plows.

In this study, the air drying of cumin seeds was boosted by cold plasma pre-treatment (CPt) followed by high-power ultrasound waves (USp). To examine the impact of included effects, different CP exposure times (0, 15, and 30 s), sonication powers (0, 60, 120, and 180 W), and drying air temperatures (30, 35, and 40 ºC) were selected as input variables. A series of well-designed experiments were conducted to evaluate drying time, effective moisture diffusivity, and energy consumption, as well as color change and rupture force of dried seeds for each drying program. Numerical investigations can effectively bypass the challenges associated with experimental analysis. Therefore, the wavelet-based neural network (WNN), the multilayer perceptron neural network (MLPNN), and the radial-basis function neural network (RBFNN), as three well-known artificial neural networks models, were used to map the inputs and output data and the results were compared with the Multiple Quadratic Regression (MQR) analysis. According to the results, the WNN model with an average correlation coefficient of R 2 > 0.92 for the train data set, and R 2 > 0.83 for the test data set provided the most beneficial tool for evaluating the drying process of cumin seeds.

Feedforward Neural Network (FNN)-based language models estimate the probability of the next word based on the history of the last N words, whereas Recurrent Neural Networks (RNN) perform the same task based only on the last word and some context information that cycles in the network. This paper presents a novel approach, which bridges the gap between these two categories of networks. In particular, we propose an architecture which takes advantage of the explicit, sequential enumeration of the word history in FNN structure while enhancing each word representation at the projection layer through recurrent context information that evolves in the network. The context integration is performed using an additional word-dependent weight matrix that is also learned during the training. Extensive experiments conducted on the Penn Treebank (PTB) and the Large Text Compression Benchmark (LTCB) corpus showed a significant reduction of the perplexity when compared to state-of-the-art feedforward as well as recurrent neural network architectures.

The purposes of this research were: (1) validating Kim's (2007) simulation method by applying analytic methods and (2) comparing the two different Robust Parameter Design methods with three measures of performance (label accuracy for enemy, friendly, and clutter). Considering the features of CID, input variables were defined as two controllable (threshold combination of detector and classifier) and three uncontrollable (map size, number of enemies and friendly). The first set of experiments considers Kim's method using analytical methods. In order to create response variables, Kim's method uses Monte Carlo simulation. The output results showed no difference between simulation and the analytic method. The second set of experiments compared the measures of performance between a standard RPD used by Kim and a new method using Artificial Neural Networks (ANNs). To find optimal combinations of detection and classification thresholds, Kim's model uses regression with a combined array design, whereas the ANNs method uses ANN with a crossed array design. In the case of label accuracy for enemy, Kim's solution showed the higher expected value, however it also showed a higher variance. Additionally, the model's residuals were higher for Kim's model.

Massive small unmanned aerial vehicles are envisioned to operate in the near future. While there are lots of research problems need to be addressed before dense operations can happen, trajectory modeling remains as one of the keys to understand and develop policies, regulations, and requirements for safe and efficient unmanned aerial vehicle operations. The fidelity requirement of a small unmanned vehicle trajectory model is high because these vehicles are sensitive to winds due to their small size and low operational altitude. Both vehicle control systems and dynamic models are needed for trajectory modeling, which makes the modeling a great challenge, especially considering the fact that manufactures are not willing to share their control systems. This work proposed to use a neural network approach for modelling small unmanned vehicle's trajectory without knowing its control system and bypassing exhaustive efforts for aerodynamic parameter identification. As a proof of concept, instead of collecting data from flight tests, this work used the trajectory data generated by a mathematical vehicle model for training and testing the neural network. The results showed great promise because the trained neural network can predict 4D trajectories accurately, and prediction errors were less than 2:0 meters in both temporal and spatial dimensions.

Neural network relationships between the full-scale, flight test hub accelerations and the corresponding three N/rev pilot floor vibration components (vertical, lateral, and longitudinal) are studied. The present quantitative effort on the UH-60A Black Hawk hub accelerations considers the lateral and longitudinal vibrations. An earlier study had considered the vertical vibration. The NASA/Army UH-60A Airloads Program flight test database is used. A physics based maneuver-effect- factor (MEF), derived using the roll-angle and the pitch-rate, is used. Fundamentally, the lateral vibration data show high vibration levels (up to 0.3 g's) at low airspeeds (for example, during landing flares) and at high airspeeds (for example, during turns). The results show that the advance ratio and the gross weight together can predict the vertical and the longitudinal vibration. However, the advance ratio and the gross weight together cannot predict the lateral vibration. The hub accelerations and the advance ratio can be used to satisfactorily predict the vertical, lateral, and longitudinal vibration. The present study shows that neural network based representations of all three UH-60A pilot floor vibration components (vertical, lateral, and longitudinal) can be obtained using the hub accelerations along with the gross weight and the advance ratio. The hub accelerations are clearly a factor in determining the pilot vibration. The present conclusions potentially allow for the identification of neural network relationships between the experimental hub accelerations obtained from wind tunnel testing and the experimental pilot vibration data obtained from flight testing. A successful establishment of the above neural network based link between the wind tunnel hub accelerations and the flight test vibration data can increase the value of wind tunnel testing.

The overall objective of this ongoing effort is to provide the capability to model and simulate rotorcraft aeromechanics behaviors in real-time. This would be accomplished by the addition of an aeromechanics element to an existing high-fidelity, real-time helicopter flight simulation. As a first step, the peak vertical vibration at the pilot floor location was considered in this neural-network-based study. The flight conditions considered were level flights, rolls, pushovers, pull-ups, autorotations, and landing flares. The NASA/Army UH-60A Airloads Program flight test database was the source of raw data. The present neural network training databases were created in a physically consistent manner. Two modeling approaches, with different physical assumptions, were considered. The first approach involved a maneuver load factor that was derived using the roll-angle and the pitch-rate. The second approach involved the three pilot control stick positions. The resulting, trained back-propagation neural networks were small, implying rapid execution. The present neural-network-based approach involving the peak pilot vibration was utilized in a quasi-static manner to simulate an extreme, time-varying pull-up maneuver. For the above pull-up maneuver, the maneuver load factor approach was better for real-time simulation, i.e., produced greater fidelity, as compared to the control stick positions approach. Thus, neural networks show promise for use in high-fidelity, real-time modeling of rotorcraft vibration.

The convolutional neural networks (CNNs) have proven to be a powerful tool for discriminative learning. Recently researchers have also started to show interest in the generative aspects of CNNs in order to gain a deeper understanding of what they have learned and how to further improve them. This paper investigates generative modeling of CNNs. The main contributions include: (1) We construct a generative model for the CNN in the form of exponential tilting of a reference distribution. (2) We propose a generative gradient for pre-training CNNs by a non-parametric importance sampling scheme, which is fundamentally different from the commonly used discriminative gradient, and yet has the same computational architecture and cost as the latter. (3) We propose a generative visualization method for the CNNs by sampling from an explicit parametric image distribution. The proposed visualization method can directly draw synthetic samples for any given node in a trained CNN by the Hamiltonian Monte Carlo (HMC) algorithm, without resorting to any extra hold-out images. Experiments on the challenging ImageNet benchmark show that the proposed generative gradient pre-training consistently helps improve the performances of CNNs, and the proposed generative visualization method generates meaningful and varied samples of synthetic images from a large-scale deep CNN.

Here we demonstrate that the activity of neural ensembles can be quantitatively modeled. We first show that an ensemble dynamical model (EDM) accurately approximates the distribution of voltages and average firing rate per neuron of a population of simulated integrate-and-fire neurons. EDMs are high-dimensional nonlinear dynamical models. To faciliate the estimation of their parameters we present a dimensionality reduction method and study its performance with simulated data. We then introduce and evaluate a maximum-likelihood method to estimate connectivity parameters in networks of EDMS. Finally, we show that this model an methods accurately approximate the high-gamma power evoked by pure tones in the auditory cortex of rodents. Overall, this article demonstrates that quantitatively modeling brain activity at the ensemble level is indeed possible, and opens the way to understanding the computations performed by neural ensembles, which could revolutionarize our understanding of brain function.

Introduction Garlic ( Allium sativum L.) is an important Allium crop in the world. Due to its therapeutic properties, it was cultivated in many countries. Furthermore, garlic is usually used as a flavoring agent; it may be used in the shape of powder or granule as a valuable condiment for foods. In addition to its use in food products, it was also widely used as an anticancer agent. Shallot ( Allium hiertifolium Boiss. L) is a perennial and bulbous plant. It is from Alliaceae family and is an important medicinal plant. The shallot is native of Iran, and grows in the high pastures. Shallot is consumed in dry areas in most parts of the country. Also shallots have been well known in Iranian folk medicine and its bulbs have been widely used for treating rheumatic and inflammatory disorders. In addition, this plant is used in the preparation of significant amounts of potassium, phosphorus, calcium, magnesium, sodium, pickles and as an additive to yogurt and pickles. ANN as a modern approach has successfully been used to solve an extensive variety of problems in the science and engineering, exclusively for some space where the conventional modeling procedure fail. A well-trained ANN can be used as a predictive model for a special use, which is a data processing system inspired by biological neural system. When mathematical equations are difficult to extrapolate, and fuzzy logic is better when decisions must be made with the estimated values below the incomplete information. The fuzzy logic theory effectively addresses the uncertainty problems that solve the ambiguity. Materials and Methods  The aim of this study was to predict moisture ratio of garlic and shallot during the drying process with fluidized bed dryer using mathematical model, artificial neural networks and fuzzy logic methods. Tests were carried out on three levels of inlet air temperature (40, 55 and 70 °C) and three inlet air velocities (0.5, 1.5 and 2.5 m s -1 ). To estimate the drying kinetic of garlic and shallot, five mathematical models were used to fit the experimental data of thin layer drying. Three factors (air temperature, air velocity and drying time) to forecast moisture ratio in fluidized bed dryer as independent variables for artificial neural networks and fuzzy logic was considered. Cascade forward back propagation (CFBP) and feed forward back propagation (FFBP) with Levenberg-Marquardt (LM), Bayesian learning (BR) algorithms for ANN and the Mamdani Fuzzy Inference System using triangular membership function were used for training patterns. Results and Discussion Consequently, the Page and Midilli et al. model was selected as the best mathematical model to describe the drying kinetics of the garlic and shallot slices, respectively. The results of artificial neural networks model for predicting MR showed that the R 2 of 0.9994 and 0.9996; and and RMSE of 0.0036 and 0.0014 were obtained for garlic and shallot, respectively. Also, The fuzzy inference system presented the R 2 of 0.9997 and 0.9998; and and RMSE of 0.0027 and 0.0011 for garlic and shallot, respectively. Comparing the results obtained from mathematical models, artificial neural networks and fuzzy logic, showed that the RMSE in the fuzzy logic was lower than artificial neural network and mathematical models. Conclusion Three factors (air temperature, air velocity and drying time) were considered for forecasting moisture ratio in fluidized bed dryer as independent variables using mathematical model, artificial neural networks and fuzzy logic. Cascade forward back propagation (CFBP) and feed forward back propagation (FFBP) with Levenberg-Marquardt (LM), Bayesian learning (BR) algorithms and the Mamdani Fuzzy Inference System using triangular membership function were used for training the patterns. Comparing the results obtained from mathematical models, artificial neural networks and fuzzy logic, showed that the root mean square error in fuzzy logic was lower than others.

Large amount of small Unmanned Aerial Vehicles (sUAVs) are projected to operate in the near future. Potential sUAV applications include, but not limited to, search and rescue, inspection and surveillance, aerial photography and video, precision agriculture, and parcel delivery. sUAVs are expected to operate in the uncontrolled Class G airspace, which is at or below 500 feet above ground level (AGL), where many static and dynamic constraints exist, such as ground properties and terrains, restricted areas, various winds, manned helicopters, and conflict avoidance among sUAVs. How to enable safe, efficient, and massive sUAV operations at the low altitude airspace remains a great challenge. NASA's Unmanned aircraft system Traffic Management (UTM) research initiative works on establishing infrastructure and developing policies, requirement, and rules to enable safe and efficient sUAVs' operations. To achieve this goal, it is important to gain insights of future UTM traffic operations through simulations, where the accurate trajectory model plays an extremely important role. On the other hand, like what happens in current aviation development, trajectory modeling should also serve as the foundation for any advanced concepts and tools in UTM. Accurate models of sUAV dynamics and control systems are very important considering the requirement of the meter level precision in UTM operations. The vehicle dynamics are relatively easy to derive and model, however, vehicle control systems remain unknown as they are usually kept by manufactures as a part of intellectual properties. That brings challenges to trajectory modeling for sUAVs. How to model the vehicle's trajectories with unknown control system? This work proposes to use a neural network to model a vehicle's trajectory. The neural network is first trained to learn the vehicle's responses at numerous conditions. Once being fully trained, given current vehicle states, winds, and desired future trajectory, the neural network should be able to predict the vehicle's future states at next time step. A complete 4-D trajectory are then generated step by step using the trained neural network. Experiments in this work show that the neural network can approximate the sUAV's model and predict the trajectory accurately.

During the past decade, machine learning has become extremely popular and can be found in many aspects of our every day life. Nowayadays with explosion of data while rapid growth of computation capacity, Distributed Deep Neural Networks (DDNNs) which can improve their performance linearly with more computation resources, have become hot and trending. However, there has not been an in depth study of the performance of these systems, and how well they scale. In this paper we analyze CNTK, one of the most commonly used DDNNs, by first building a performance model and then evaluating the system two settings: a small cluster with all nodes in a single rack connected to a top of rack switch, and in large scale using Blue Waters with arbitary placement of nodes. Our main focus was the scalability of the system with respect to adding more nodes. Based on our results, this system has an excessive initialization overhead because of poor I/O utilization which dominates the whole execution time. Because of this, the system does not scale beyond a few nodes (4 in Blue Waters). Additionally, due to a single server-multiple worker design the server becomes a bottleneck after 16 nodes limiting the scalability of the CNTK.

Speaker intent detection and semantic slot filling are two critical tasks in spoken language understanding (SLU) for dialogue systems. In this paper, we describe a recurrent neural network (RNN) model that jointly performs intent detection, slot filling, and language modeling. The neural network model keeps updating the intent estimation as word in the transcribed utterance arrives and uses it as contextual features in the joint model. Evaluation of the language model and online SLU model is made on the ATIS benchmarking data set. On language modeling task, our joint model achieves 11.8% relative reduction on perplexity comparing to the independent training language model. On SLU tasks, our joint model outperforms the independent task training model by 22.3% on intent detection error rate, with slight degradation on slot filling F1 score. The joint model also shows advantageous performance in the realistic ASR settings with noisy speech input.

When humans read text, they fixate some words and skip others. However, there have been few attempts to explain skipping behavior with computational models, as most existing work has focused on predicting reading times (e.g.,~using surprisal). In this paper, we propose a novel approach that models both skipping and reading, using an unsupervised architecture that combines a neural attention with autoencoding, trained on raw text using reinforcement learning. Our model explains human reading behavior as a tradeoff between precision of language understanding (encoding the input accurately) and economy of attention (fixating as few words as possible). We evaluate the model on the Dundee eye-tracking corpus, showing that it accurately predicts skipping behavior and reading times, is competitive with surprisal, and captures known qualitative features of human reading.

In this study, it was aimed to predict elementary education teacher candidates' achievements in "Science and Technology Education I and II" courses by using artificial neural networks. It was also aimed to show the independent variables importance in the prediction. In the data set used in this study, variables of gender, type of education, field of study in high school and transcript information of 14 courses including end-of-term letter grades were collected. The fact that the artificial neural network performance in this study was R = 0.84 for the Science and Technology Education I course, and R = 0.84 for the Science and Technology Education II course shows that the network performance overlaps with the findings obtained from the related studies.

This study is aimed to explore using an artificial neural network method to anticipate the confined compressive strength and its corresponding strain for the circular concrete columns wrapped with FRP sheets. 58 experimental data of circular concrete columns tested under concentric loading were collected from the literature. The experimental data is used to train and test the neural network. A comparative study was also carried out between the neural network model and the other existing models. It was found that the fundamental behavior of confined concrete columns can logically be captured by the neural network model. Besides, the neural network approach provided better results than the analytical and experimental models. The neural network-based model with R 2  equal to 0.993 and 0.991 for training and testing the compressive strength, respectively, shows that the presented model is a practical method to predict the confinement behavior of concrete columns wrapped with FRP since it provides instantaneous result once it is appropriately trained and tested.

This study is aimed to explore using an artificial neural network method to anticipate the confined compressive strength and its corresponding strain for the circular concrete columns wrapped with FRP sheets. 58 experimental data of circular concrete columns tested under concentric loading were collected from the literature. The experimental data is used to train and test the neural network. A comparative study was also carried out between the neural network model and the other existing models. It was found that the fundamental behavior of confined concrete columns can logically be captured by the neural network model. Besides, the neural network approach provided better results than the analytical and experimental models. The neural network-based model with R 2  equal to 0.993 and 0.991 for training and testing the compressive strength, respectively, shows that the presented model is a practical method to predict the confinement behavior of concrete columns wrapped with FRP since it provides instantaneous result once it is appropriately trained and tested.

Click here to view the University of Florida catalog record

Aircraft design and control techniques rely on the proper modeling of the aircraft's equations of motion. Many of the variables used in these equations are aerodynamic coefficients which are obtained from scale models in wind tunnel tests. In order to model damaged aircraft, every aerodynamic coefficient must be determined for every possible damage mechanism in every flight condition. Designing a controller for a damaged aircraft is particularly burdensome because knowledge of the effect of each damage mechanism on the model is required before the controller can be designed. Also, a monitoring system must be employed to decide when and how much damage has occurred in order to re configure the controller. Recent advances in artificial intelligence have made parallel distributed processors (artificial neural networks) feasible. Modeled on the human brain, the artificial neural network's strength lies in its ability to generalize from a given model. This thesis examines the robustness of the artificial neural network as a model for damaged aircraft.

In this paper, we propose an efficient transfer leaning methods for training a personalized language model using a recurrent neural network with long short-term memory architecture. With our proposed fast transfer learning schemes, a general language model is updated to a personalized language model with a small amount of user data and a limited computing resource. These methods are especially useful for a mobile device environment while the data is prevented from transferring out of the device for privacy purposes. Through experiments on dialogue data in a drama, it is verified that our transfer learning methods have successfully generated the personalized language model, whose output is more similar to the personal language style in both qualitative and quantitative aspects.

A central challenge in neuroscience is understanding how neural system implements computation through its dynamics. We propose a nonlinear time series model aimed at characterizing interpretable dynamics from neural trajectories. Our model assumes low-dimensional continuous dynamics in a finite volume. It incorporates a prior assumption about globally contractional dynamics to avoid overly enthusiastic extrapolation outside of the support of observed trajectories. We show that our model can recover qualitative features of the phase portrait such as attractors, slow points, and bifurcations, while also producing reliable long-term future predictions in a variety of dynamical models and in real neural data.

Neural network technology has recently demonstrated capabilities in areas important to personnel research such as statistical analysis, decision modeling, control, and forecasting. The present investigation indicates that three different neural network architectures are particularly suited to modeling many aspects of the Air Force personnel system: back propagation, learning vector quantization, and probabilistic neural networks. The primary advantage of neutral networks is their ability to derive nonlinear and interacting relationships among model variables. Two areas investigated in order to evaluate this capability were airmen reenlistment decisions and airman inventory modeling.

In this work, baseline energy model development using Artificial Neural Network (ANN) with resampling techniques; Cross Validation (CV) and Bootstrap (BS) are presented. Resampling techniques are used to examine the ability of the ANN model to deal with a small dataset. Working days, class days and Cooling Degree Days (CDD) are used as ANN input meanwhile the ANN output is monthly electricity consumption. The coefficient of correlation (R) is used as performance function to evaluate the model accuracy. For this analysis, R is calculated for the entire data set (R_all) and separately for training set (R_train), validation set (R_valid) dan testing set (R_test). The closer R to 1, the higher similarities between targeted and predicted output. The total of two different models with several number of neurons are developed and compared. It can be concluded that all models are capable to train the network. Artificial Neural Network with Bootstrap Cross Validation technique (ANN-BSCV) outperforms Artificial Neural Network with Cross Validation technique (ANN-CV). The 3-6-1 ANN-BSCV, with R_train = 0.95668, R_valid = 0.97553, R_test = 0.85726 and R_all = 0.94079 is selected as the baseline energy model to predict energy consumption for Option C IPMVP. 

Introduction Soil compaction is one of the most destructive effects of machine traffic. Compaction increases soil mechanical strength and reduces its porosity, plant rooting and ultimately the yield. Nowadays, agricultural machinery has the maximum share on soil compaction in modern agriculture. The soil destruction may be as surface deformation or as subsurface compaction. Any way machine traffic destructs soil structure and as result has unfavorable effect on the yield. Hence, soil compaction recognition and its management are very important. In general, soil compaction is the most destructive effect of machine traffic. Modeling of ecological systems by conventional modeling methods due to the multitude effective parameters has always been challenging. Artificial intelligence and soft computing methods due to their simplicity, high precision in simulation of complex and nonlinear processes are highly regarded. The purpose of this research was the modeling of soil compaction system affected by soil moisture content, the tractor forward velocity and soil depth by multilayer perceptron neural network. Materials and Methods In order to carry out the field experiments, a tractor MF285 which was equipped with a three-tilt moldboard plough was used. Experiments were conducted at the Agricultural research field of University of Mohaghegh Ardabili in five levels of moisture content of 11, 14, 16, 19 and 22%, forward velocity of 1, 2, 3, 4 and 5 km.h -1 , and soil depths of 20, 25, 30, 35 and 40 cm as a randomized complete block design with three replications. In this study, perceptron neural network with five neurons in the hidden layer with sigmoid transfer function and linear transfer function for the output neuron was designed and trained. Results and Discussion Field experiments showed three main factors were significant on the bulk density (P<0.01). The mutual effect of moisture on depth and mutual binary effect of moisture on velocity and depth on velocity were significant (P<0.01). Mutual triplet effect of moisture on velocity on depth was significant (P<0.05). Maximum bulk density of 1362 kg/m 3 was obtained at the highest moisture of 22% and the lowest forward velocity of 1 km/h at the depth of 20 cm. Whilst the minimum value of 1234.5 kg/m 3 was related to the moisture, forward velocity and depth of 11%, 5 km/h and depth of 40 cm, respectively. Compaction increased as soil moisture content increased up to 22% which was critical moisture. In contrast, soil compaction decreased as the tractor velocity and soil depth increased. A comparison of neural network output and experimental results indicated a high determination coefficient of R 2 = 0.99 between them. Also, the mean square error of the model was 0.174, in addition, mean absolute percentage error of the system (MAPE) was equal to %0.29 which showed high accuracy of neural network to model soil compaction. Conclusion It was concluded that soil compaction increased as soil moisture content increased up to a critical value. Increasing soil moisture act as lubricant and soil layers compacted together. Hence knowledge of soil moisture can help us to manage soil compaction during agricultural operations. Increasing the tractor forward velocity reduced soil compaction. However, agricultural operations should be conducted at certain speeds to carry out the duty properly and increasing speed more that value decreases the efficiency of work. Neural network of MLP with 5 neurons in hidden layer and sigmoid function in middle layer and one neuron with linear transfer function was found the most accurate and precise in prediction of the soil bulk density.  A high determination coefficient of R 2 = 0.99 was found between measured and predicted values.

Research works are ongoing in mixing the biologically synthesized oil with the diesel for reducing the effect of global warming and climate change. From the review study, it is noted that the blended biodiesels require more assert about their practical viability. So, the non-edible Tamanu oil is synthesized and it is blended with diesel and its emission characteristics, engine performance and combustion characteristics are studied in our previous work. This paper attempts to model the diesel engine fueled with tamanu oil biodiesel blend. The proposed model exploits the context of neural network and the firefly algorithm is used to train it. After analyzing the various characteristics of the diesel engine, the acquired data is subjected to a proposed FF-NM method. The simulated results are statistically evaluated and the proposed modeling method is proved to be better than the other NM.

Recently neural networks have been applied in antenna modeling where the role of the network is not just for black-box modeling. This paper highlights that aspect of neural networks from the antenna engineering application point of view.

We investigate the theory of observers in the quantum mechanical world by using a novel model of the human brain which incorporates the glial network into the Hopfield model of the neural network. Our model is based on a microscopic construction of a quantum Hamiltonian of the synaptic junctions. Using the Eguchi-Kawai large N reduction, we show that, when the number of neurons and astrocytes is exponentially large, the degrees of freedom of the dynamics of the neural and glial networks can be completely removed and, consequently, that the retention time of the superposition of the wave functions in the brain is as long as that of the microscopic quantum system of pre-synaptics sites. Based on this model, the classical information entropy of the neural-glial network is introduced. Using this quantity, we propose a criterion for the brain to be a quantum mechanical observer.

How to model a pair of sentences is a critical issue in many NLP tasks such as answer selection (AS), paraphrase identification (PI) and textual entailment (TE). Most prior work (i) deals with one individual task by fine-tuning a specific system; (ii) models each sentence's representation separately, rarely considering the impact of the other sentence; or (iii) relies fully on manually designed, task-specific linguistic features. This work presents a general Attention Based Convolutional Neural Network (ABCNN) for modeling a pair of sentences. We make three contributions. (i) ABCNN can be applied to a wide variety of tasks that require modeling of sentence pairs. (ii) We propose three attention schemes that integrate mutual influence between sentences into CNN; thus, the representation of each sentence takes into consideration its counterpart. These interdependent sentence pair representations are more powerful than isolated sentence representations. (iii) ABCNN achieves state-of-the-art performance on AS, PI and TE tasks.

This paper proposes a new data-driven approach for modeling detailed splashes for liquid simulations with neural networks. Our model learns to generate small-scale splash detail for fluid-implicit-particle methods using training data acquired from physically accurate, high-resolution simulations. We use neural networks to model the regression of splash formation using a classifier together with a velocity modification term. More specifically, we employ a heteroscedastic model for the velocity updates. Our simulation results demonstrate that our model significantly improves visual fidelity with a large amount of realistic droplet formation and yields splash detail much more efficiently than finer discretizations. We show this for two different spatial scales and simulation setups.

A variety of real-world processes (over networks) produce sequences of data whose complex temporal dynamics need to be studied. More especially, the event timestamps can carry important information about the underlying network dynamics, which otherwise are not available from the time-series evenly sampled from continuous signals. Moreover, in most complex processes, event sequences and evenly-sampled times series data can interact with each other, which renders joint modeling of those two sources of data necessary. To tackle the above problems, in this paper, we utilize the rich framework of (temporal) point processes to model event data and timely update its intensity function by the synergic twin Recurrent Neural Networks (RNNs). In the proposed architecture, the intensity function is synergistically modulated by one RNN with asynchronous events as input and another RNN with time series as input. Furthermore, to enhance the interpretability of the model, the attention mechanism for the neural point process is introduced. The whole model with event type and timestamp prediction output layers can be trained end-to-end and allows a black-box treatment for modeling the intensity. We substantiate the superiority of our model in synthetic data and three real-world benchmark datasets.

Water quality prediction is aided by environmental monitoring, ecological sustainability, and aquaculture. Traditional prediction approaches capture the nonlinearity and non-stationarity of water quality well. Due to their rapid progress, artificial neural networks (ANNs) have become a hotspot in water quality prediction in recent years. ANNs are utilised in this study to predict water quality using soft computing techniques. The feedforward network and the standard back-propagation method of Levenberg-Marquardt and scaled conjugate gradient learning algorithm were employed in this research. One hidden layer has been recommended for the modelling, with the number of hidden neurons set at 3, 24, and 49. For this analysis, six different testing percentages were used, and the output data can be categorised as '0' for clean water and '1' for polluted water. From the results, it can be shown that the most optimised model was from the model of trainlm with a testing percentage of 18% and with 3 number of neurons. This most optimised model obtains an accuracy of 91.7%, the best validation performance of 0.073346 with 24 epochs, and having a receiver operating characteristic (ROC) curve that is closer to the true positive rate compared to other samples.

We introduce a nonlinear modification of the classical Hawkes process, which allows inhibitory couplings between units without restrictions. The resulting system of interacting point processes provides a useful mathematical model for recurrent networks of spiking neurons with exponential transfer functions. The expected rates of all neurons in the network are approximated by a first-order differential system. We study the stability of the solutions of this equation, and use the new formalism to implement a winner-takes-all network that operates robustly for a wide range of parameters. Finally, we discuss relations with the generalised linear model that is widely used for the analysis of spike trains.

We introduce a nonlinear modification of the classical Hawkes process, which allows inhibitory couplings between units without restrictions. The resulting system of interacting point processes provides a useful mathematical model for recurrent networks of spiking neurons with exponential transfer functions. The expected rates of all neurons in the network are approximated by a first-order differential system. We study the stability of the solutions of this equation, and use the new formalism to implement a winner-takes-all network that operates robustly for a wide range of parameters. Finally, we discuss relations with the generalised linear model that is widely used for the analysis of spike trains.

This article is from Frontiers in Computational Neuroscience , volume 8 . Abstract None

This study investigates how victim attitude responses (neutral vs. negative) influence third-party punishment decisions, using EEG and computational modeling. The goal is to understand the cognitive and neural mechanisms that underlie third-party punishment when victim feedback is incorporated. The EEG experiment aims to clarify the neural indicators of different victim attitudes (neutral vs. negative) under fair and unfair conditions, while the behavioral replication experiment seeks to replicate the behavioral results observed in the EEG study. Importantly, the study combines utility models to explore how parameters change in different attitude contexts, providing insights into the underlying psychological mechanisms.

In this work, we propose contextual language models that incorporate dialog level discourse information into language modeling. Previous works on contextual language model treat preceding utterances as a sequence of inputs, without considering dialog interactions. We design recurrent neural network (RNN) based contextual language models that specially track the interactions between speakers in a dialog. Experiment results on Switchboard Dialog Act Corpus show that the proposed model outperforms conventional single turn based RNN language model by 3.3% on perplexity. The proposed models also demonstrate advantageous performance over other competitive contextual language models.

The negative sampling (NEG) objective function, used in word2vec, is a simplification of the Noise Contrastive Estimation (NCE) method. NEG was found to be highly effective in learning continuous word representations. However, unlike NCE, it was considered inapplicable for the purpose of learning the parameters of a language model. In this study, we refute this assertion by providing a principled derivation for NEG-based language modeling, founded on a novel analysis of a low-dimensional approximation of the matrix of pointwise mutual information between the contexts and the predicted words. The obtained language modeling is closely related to NCE language models but is based on a simplified objective function. We thus provide a unified formulation for two main language processing tasks, namely word embedding and language modeling, based on the NEG objective function. Experimental results on two popular language modeling benchmarks show comparable perplexity results, with a small advantage to NEG over NCE.

This article is from BMC Neuroscience , volume 15 . Abstract None