Optical Memory and Neural Networks最新文献

A vital source of nutrition and a major contributor to the nation’s economic expansion is agriculture. Due to numerous complex factors such as environment, humidity, soil nutrients, and soil moisture, multi crop yield forecasting was very challenging. Because crop prediction is a complicated process, improving performance is challenging. To address these problems, an advance deep learning model was developed to predict crop types and its yields in a particular soil. A real time data were created, which contain various parameters such as soil nutrition’s, weather, data, seasons and temperature. The created dataset is pre-processed using outlier detection as well as normalization because it contains unwanted rows and columns. After that, the pre-processed data were given as input for the DeepNet230 model to analyze the input parameters like soil nutrition and temperature to predict the multi crop type and its yield quantity. DeepNet230 have the capacity of automatic feature learning and rapid unstructured process, so it provides an efficient prediction performance of crop yield and its types. The performance analysis of crop prediction for the proposed model are 93.7% accuracy, 93.4% recall, 92.8% precision and 92.9% specificity. Then, the performance of yield prediction for the identified crops are 95.5% accuracy, 91.6% recall, 93% precision and 94.2% specificity. In addition, the developed method was compared with several opposing methods for validation. The observed results show that the suggested method performed significantly better in real time due to its improved predictive capabilities.

Motor speech condition called dysarthria is caused by a lack of movement in the lips, tongue, vocal cords, and diaphragm are a few of the muscles needed to produce speech. Speech that is slurred, sluggish, or inaccurate might be the initial sign of dysarthria, which varies in severity. Parkinson’s disease, muscular dystrophy, multiple sclerosis, brain tumors, brain damage, and amyotrophic lateral sclerosis are among the health problems that can result from dysarthria. This research develops an efficient method for extracting features and classifying dysarthria affected persons from speech signals. This suggested method uses a speech signal as its source. The supplied speech signal is pre-processed to improve the identification of dysarthria speech. Pre-processing methods like the Butterworth band pass filter and Savitzky Golay digital FIR filter are used to smoothing the raw data. After pre-processing, the signals are input into the feature extraction techniques, such as Yule-Walker Autoregressive modelling, Mel frequency cepstral coefficients and Perceptual Linear Predictive to extract the important features. The dysarthria speech is finally detected using an improved Elman Spike Neural Network (EESNN) algorithm-based classifier. Hunter Prey Optimization (HPO) is used to select the weights of EESNN optimally. The proposed algorithm achieves 94.25% accuracy and 94.26% specificity values. Thus this proposed approach is the best choice for predicting dysarthria disease using speech signal.

In this paper proposes an algorithm for the analytical calculation of convolutional neural networks without using neural network training algorithms. A description of the algorithm is given, on the basis of which the weights and threshold values of a convolutional neural network are analytically calculated. In this case, to calculate the parameters of the convolutional neural network, only 10 selected samples were used from the MNIST digit database, each of which is an image of one of the recognizable classes of digits from 0 to 9, and was randomly selected from the MNIST digit database. As a result of the operation of this algorithm, the number of channels of the convolutional neural network layers is also determined analytically. Based on the proposed algorithm, a software module was implemented in the Builder environment C++, on the basis of which a number of experiments were carried out with recognition of the MNIST database. The results of the experiments described in the work showed that the computation time of convolutional neural networks is very short and amounts to fractions of a second or a minute. Analytically computed convolutional neural networks were tested on the MNIST digit database, consisting of 1000 images of handwritten digits. The experimental results showed that already using only 10 selected images from the MNIST database, analytically calculated convolutional neural networks are able to recognize more than half of the images of the MNIST database, without application of neural network training algorithms. In general, the study showed that artificial neural networks, and in particular convolutional neural networks, are capable of not only being trained by learning algorithms, but also recognizing images almost instantly, without the use of learning algorithms using preliminary analytical calculation of the values of the neural network’s weights.

The paper elaborates on methods of digital image compression. The focus is on the compression method that represents a raster image as a set of multiply thinned sub-images. Sub-images are processed consecutively to generate special reference images. The difference between the synthesized reference image and original sub-image forms a divergence array. The algorithm introduces a discrete error into the divergence array to provide the actual bit-depth reduction. However, the introduction of the error inevitably impairs the quality of the decompressed image. The aim is to make sure that the parametric smoothing of divergence arrays can lessen this quality impairment without changing the bit depth reduction originally provided by the method. Numerical experiments on real digital images are carried out to prove that the use of parametric smoothing improves noticeably the efficiency of the image compression method under discussion.

The clean hydrogen is needed for green energy. It can be obtained by the water electrolysis, which is energetically unprofitable. The problem of hydrogen storage solution made it possible to use it as an automobile fuel. There was a place for the laser in the cramped fuel cell. Previously, it was proposed to introduce laser radiation with the wavelengths corresponding to the water molecule vibrational levels excitation into the reaction zone to increase energy efficiency. In addition, all processes on the Earth should be considered taking into account hydrogen degassing, that is, the hydrogen escape from the Earth into the atmosphere. And so the laser is the most suitable tool for finding places where the hydrogen exits to the surface. In this paper, it is proposed to use the Raman lidar for laser remote sensing of the hydrogen molecules during its leaks into the atmosphere. Based on the results of the Raman lidar equation computer simulation in the range of ranging distances up to 100 m, it is shown that its parameters optimization will reduce the values of detectable concentrations of the hydrogen molecules in the atmosphere.

In Ethernet system, as a result of head of line blocking, numerous control data queues with high priority may cause priority queues to become overcrowded and their receiving DMAs (Direct Memory Access) to run out of buffer space, forcing them to delete packets that are still arriving from the network. Thus the primary goal of this work is to prevent deadlock in an Ethernet system while sending congested information across the Ethernet protocol and channel. In order to allow many processors to interact concurrently without causing a conflict, this research paper proposes a Memory crossbar switching control in which the memory is divided into global and local partitions utilizing the q-learning architecture in the development of a Q-Memory task routing architecture. The path average value therefore represents congestion information for each router and its surrounding nodes. The nearby router receives the path average value if the message is received. The networks-on-chip protocol and channel should be used to provide congestion information in order to prevent deadlock in a system and improve communication.

Air pollution imposed by particle matter (PM) made it a public health concern and hazard to humans and the environment. Reduced vision, allergic responses, pneumonia, asthma, cardiovascular disorders, lung cancer, and even mortality can result from prolonged exposure to the concentration of air’s small particulate matter. Air quality prediction can offer reliable information for future air pollution status to operate air pollution control effectively and make preventative plans. Tracking, predicting, and regulating emissions is crucial. Controlling PM2.5 is the key for enhancing air quality, and it can be accomplished by forecasting PM2.5 concentrations. This work develops a methodology for forecasting PM2.5 concentrations using a gradient-boosted regression tree with Convolutional Neural Network (CNN) and fuzzy K-nearest neighbour (fuzzy-KNN). The results of the proposed methodology have been comparatively analysed with multiple linear regression, stacked long short-term memory, bidirectional gated recurrent unit, and gradient-boosted regression tree. The Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) are evaluated, and it shows that the gradient-boosted regression tree model produces a reduced error with improved accuracy in forecasting air quality.

Most of the existing fake news identification models mainly focused on exploiting multi-modal features to enhanced performance recently. This paper proposes Multi-modal Co-Attention Capsules Network (MCCN) for fake news detection, which consists mainly of feature extraction layer, feature fusion layer and classification layer. Feature extraction layer achieves the features building of users’ profiles, multi-modal source news and comments. Feature fusion layer adopts a dual parallel Cross-Modal Co-Attentional to fuse multi-modal interactions between source news text and its attached image, Hierarchical Co-Attention to fuse the interactions among user information, source news content and comments. Classification layer adopts capsules network to realize false information identification. Experimental results on three widely used large-scale datasets show that MCCN can achieve the excellent performance by comparing with other baselines.

Healthcare industry is a stage which is presented with tremendous innovative headways consistently. Parkinson disease (PD) has become a critical overall general clinical issue starting late. To provide the solution for this problem, in this paper, use fusion of machine learning and federated learning techniques for processing electronically collected patients’ health record (PD dataset) in accurate manner. The PD dataset are constantly gathered and sorted out to give a point by point history of patients, their sicknesses and determination plans. The medical PD dataset contains 43 400 electronic records of potential patients which includes normal, Ischemic and Hemorrhagic stroke. Cleaning, finding feature correlation and imputing missing values in the PD has to be performed by preprocessing & normalization approach. For further processing, using Random over sampling (ROS) methods the imbalanced PD dataset will be converted into balanced. From the balanced PD datasets the stroke prediction accuracy will be validated using Decision Tree, Logistic Regression, Random Forest and Improved LSTM (Imp-LSTM) machine learning algorithms. Using distinct experiments of executing performance measurements the accuracy rate from our prediction classifiers for the patient with smokes category will be 62.29, 71.36, 96.51 and 99.56% respectively as like the patient with never smoked category dataset the accuracy will be 70.49, 75.86, 96.49 and 99.58% respectively. The proposed Imp-LSTM algorithm in this research will effectively produce high overall accuracy in both the datasets, which means a successful decrease in the misdiagnosis rate for stroke prediction.

Machine learning is currently one of the most actively developing research areas. Considerable attention in the ongoing research is paid to problems related to dynamical systems. One of the areas in which the application of machine learning technologies is being actively explored is aircraft of various types and purposes. This state of the art is due to the complexity and variety of tasks that are assigned to aircraft. The complicating factor in this case is incomplete and inaccurate knowledge of the properties of the object under study and the conditions in which it operates. In particular, a variety of abnormal situations may occur during flight, such as equipment failures and structural damage, which must be counteracted by reconfiguring the aircraft’s control system and controls. The aircraft control system must be able to operate effectively under these conditions by promptly changing the parameters and/or structure of the control laws used. Adaptive control methods allow to satisfy this requirement. One of the ways to synthesize control laws for dynamic systems, widely used nowadays, is LQR approach. A significant limitation of this approach is the lack of adaptability of the resulting control law, which prevents its use in conditions of incomplete and inaccurate knowledge of the properties of the control object and the environment in which it operates. To overcome this limitation, it was proposed to modify the standard variant of LQR (Linear Quadratic Regulator) based on approximate dynamic programming, a special case of which is the adaptive critic design (ACD) method. For the ACD-LQR combination, the problem of controlling the lateral motion of a maneuvering aircraft is solved. The results obtained demonstrate the promising potential of this approach to controlling the airplane motion under uncertainty conditions.