Search

This study proposes a new method for diagnosing Parkinson's disease using audio signals and FPGA-based convolutional neural networks. The proposed method involves training a convolutional neural network and using deep learning techniques to increase its accuracy. The method is implemented on an FPGA chip to reduce latency and storage space. This work achieves higher accuracy and lower delay than other methods. The study also explores the potential of using the dropout approach during network training to improve the learning of convolutional neural networks. The findings of this study suggest that the proposed method can be a useful tool for diagnosing Parkinson's disease with audio data. Contributions and novelties of this work include the use of FPGA-based convolutional neural networks for Parkinson's disease diagnosis, the implementation of the method on an FPGA chip, and the exploration of the dropout approach for improving the accuracy of convolutional neural networks. The proposed method achieves higher accuracy and lower delay than other methods, making it a useful tool for diagnosing Parkinson's disease with audio data. The use of a new accelerator with a dropout technique and lowering the learning rate to increase detection accuracy is also a novel contribution to the field of machine learning. In addition, the proposed method eliminates the need to store convolution mapping, reducing the chip volume and computational volume, and the integrated chip is much smaller in volume and uses less power than prior generations. The method is developed and trained on audio data to detect Parkinson's disease and then implemented on an FPGA using a high-end hardware accelerator architecture. By reducing the bit width, this method decreases the space used by the logical cells and minimizes latency via the proposed computational model and parallel computations. [ABSTRACT FROM AUTHOR]

Currently, cloud computing provides the necessary infrastructure and software services to provide the requested services needed by users on the Internet. Due to the spectacular growth of cloud computing, the number of users and the number of demands are increasing rapidly, which creates a high workload on servers and computing resources. In this situation, for the optimal use of resources, the need for an efficient and effective management approach is fully felt. For this purpose, game theory has been used. The game structure is designed in such a way that the leader (leader) owns a large number of resources and plans the allocation of resources based on the request received from mobile users. The goal from the leader’s point of view is to minimize the cost of using the resources located in the fog nodes, on the other hand, the goal considered from the mobile users’ point of view is to minimize the cost of responding and transmitting the message to the desired fog node. For this purpose, the entire region is divided into regions and a fog node is considered for each region. The main goal of this research is to reduce the average delay in the provision of services related to Internet of Things applications in cloud computing platform. For this purpose, an attempt is made to provide a new method for allocating multiple tasks in mobile crowd sensing based on fog computing in Internet of Things using the inverse Stackelberg game theory with the help of fuzzy logic and deep reinforcement learning algorithm. [ABSTRACT FROM AUTHOR]

CsPbI3 perovskite quantum dots (CPQDs) have received great attention due to their potential in large-scale applications. Increasing the efficiency of CPQDs solar cells is an important issue that is addressed in this paper. Here, we have simulated a 14.61% colloidal CPQD solar cell with the least fitting parameter that shows the accuracy of the following results. The absorber layer properties are varied and different power conversion efficiency (PCE) is achieved for the new device. The results show that colloidal CsPbI3 material properties have a significant effect on the PCE of the device. Finally, the optimized parameters for the absorber layer are listed and the optimum efficiency of 29.88% was achieved for this case. Our results are interesting that help the researchers to work on CsPbI3 materials for the achievement of highly efficient, stable, large-scale, and flexible CPQDs solar cells. [ABSTRACT FROM AUTHOR]

CsPbI3 perovskite quantum dots (CPQDs) have received great attention due to their potential in large-scale applications. Increasing the efficiency of CPQDs solar cells is an important issue that is addressed in this paper. Here, we have simulated a 14.61% colloidal CPQD solar cell with the least fitting parameter that shows the accuracy of the following results. The absorber layer properties are varied and different power conversion efficiency (PCE) is achieved for the new device. The results show that colloidal CsPbI3 material properties have a significant effect on the PCE of the device. Finally, the optimized parameters for the absorber layer are listed and the optimum efficiency of 29.88% was achieved for this case. Our results are interesting that help the researchers to work on CsPbI3 materials for the achievement of highly efficient, stable, large-scale, and flexible CPQDs solar cells. [ABSTRACT FROM AUTHOR]

Most web-based educational systems contain some drawbacks, as compared to traditional classrooms. Particularly, it becomes difficult for teachers to guide students to choose an appropriate learning resource due to the large number of online learning resources. Meanwhile, student decisions make it more difficult to choose educational resources according to their circumstances. In this matter, the resource recommender system can be employed as an educational environment to recommend the educational resource advice for students, so that these recommendations can be coordinated to each student's preferences and needs. This paper presents the resource recommender system as a combination of MLP, BiLSTM, and LSTM improved deep learning networks using the attention method. Compared to similar studies conducted using DBN networks and focus only on the near past interests and preferences of users, the proposed system provides higher accuracy and more appropriate recommendations considering current interests, in addition to the user's long-term past interests. The proposed recommender system with accuracy of 0.96 and a loss of 0.0822 contains a better performance to recommend resources to students compared to other methods. [ABSTRACT FROM AUTHOR]

Cancer is the second important morbidity and mortality factor among women and the most incident type is breast cancer. This paper suggests a hybrid computational intelligence model based on unsupervised and supervised learning techniques, i.e., self-organizing map (SOM) and complex-valued neural network (CVNN), for reliable detection of breast cancer. The dataset used in this paper consists of 822 patients with five features (patient's breast mass shape, margin, density, patient's age, and Breast Imaging Reporting and Data System assessment). The proposed model was used for the first time and can be categorized in two stages. In the first stage, considering the input features, SOM technique was used to cluster the patients with the most similarity. Then, in the second stage, for each cluster, the patient's features were applied to complex-valued neural network and dealt with to classify breast cancer severity (benign or malign). The obtained results corresponding to each patient were compared to the medical diagnosis results using receiver operating characteristic analyses and confusion matrix. In the testing phase, health and disease detection ratios were 94 and 95%, respectively. Accordingly, the superiority of the proposed model was proved and can be used for reliable and robust detection of breast cancer. [ABSTRACT FROM AUTHOR]

In this article, a new configuration of a three‐stage Complementary metal‐oxide‐semiconductor (CMOS) operational transconductance amplifier (OTA) is proposed. It removes the feedforward path while simultaneously amplifying the feedback path in the compensation network. Also, the proposed circuit uses only one small compensation capacitor which makes it appropriate regarding die area. The approach behind is demonstrated through a successful design using 0.18‐μm CMOS technology. In fact, the simulated OTA shows better performance in comparison with other methods and achieved 20.2 MHz, 82°, and 124 dB as gain‐bandwidth product, phase margin, and DC gain, respectively, and consumes 545 μW @ 1.8 V. [ABSTRACT FROM AUTHOR]