Hyper-parameter optimization of convolutional neural network based on particle swarm optimization algorithm

Deep neural networks have accomplished enormous progress in tackling many problems. More specifically, convolutional neural network (CNN) is a category of deep networks that have been a dominant technique in computer vision tasks. Despite that these deep neural networks are highly effective; the ideal structure is still an issue that needs a lot of investigation. Deep Convolutional Neural Network model is usually designed manually by trials and repeated tests which enormously constrain its application. Many hyper-parameters of the CNN can affect the model performance. These parameters are depth of the network, numbers of convolutional layers, and numbers of kernels with their sizes. Therefore, it may be a huge challenge to design an appropriate CNN model that uses optimized hyper-parameters and reduces the reliance on manual involvement and domain expertise. In this paper, a design architecture method for CNNs is proposed by utilization of particle swarm optimization (PSO) algorithm to learn the optimal CNN hyper-parameters values. In the experiment, we used Modified National Institute of Standards and Technology (MNIST) database of handwritten digit recognition. The experiments showed that our proposed approach can find an architecture that is competitive to the state-of-the-art models with a testing error of 0.87%.

Evaluation of Microscan WalkAway for determination of ceftazidime/avibactam and ceftolozane/tazobactam susceptibility in carbapenem-resistant Gram-negative bacilli

Background: We evaluated the performance of ceftazidime/avibactam and ceftolozane/tazobactam MicroScan Neg multidrug-resistant MIC 1 NMR1 panel for clinical carbapenem-nonsusceptible Gram-negative bacilli isolates. Methods: We evaluated 212 clinically significant carbapenem-nonsusceptible Gram-negative bacilli 139 Pseudomonas aeruginosa and 73 KPC-producing Enterobacterales from 71 Brazilian hospitals 2013-2020. Ceftazidime/avibactam and ceftolozane/tazobactam MICs from the panel were compared with broth microdilution BMD test as the reference method. Essential agreement EA and categorical agreement CA were assessed. For P. aeruginosa , antimicrobial susceptibility testing error rates were calculated using the error-rate bound method. Results: Discrepancies were initially observed with 11 isolates, 4 resolved after retesting, 2 in favor of the NMR1 and 2 in favor of the BMD method. The ceftazidime/avibactam EA overall and evaluable was 100% for P. aeruginosa and Enterobacterales. CA was 100% for Enterobacterales , and 98.6% for P. aeruginosa . The ceftolozane/tazobactam EA was 98.6% and 100% overall and evaluable, respectively, and CA was 96.4% for P. aeruginosa . For ceftazidime/avibactam, no VME was found, and the ME rate was 4.2% 2/48. For ceftolozane/tazobactam and P. aeruginosa , using the CLSI breakpoints, the minor error mE was 11.4% and no VME or ME was found. While using EUCAST breakpoints, VME was 11.4% with no ME. The mE becomes ME or VME in the absence of the intermediate category. All categorical errors were also within 1 log of MIC variation, and the adjusted error rate for CLSI/EUCAST was 0% 0/212. Conclusions: The NMR1 panel is an option to test ceftazidime–avibactam for KPC-producing Enterobacterales and carbapenem-nonsusceptible P. aeruginosa . When ceftolozane–tazobactam MIC 4 mg/L are obtained using this method, an alert could be created, and the results could be confirmed by alternative method.

Error Detection in Music Education: A Review of Literature

The ability to accurately detect performance errors is a fundamental skill for music educators and has been a popular topic of research within the field of music education. In fact, it has been suggested that roughly half of all ensemble rehearsals are dedicated to error detection. The purpose of this literature review was to synthesize the research literature related to error detection among preservice and inservice music educators. The majority of error detection studies have centered on the topics of (a) defining errors and error hierarchy, (b) developing tests and programmed materials, (c) personal characteristics related to error detection ability, and (d) factors that influence error detection ability. Results from existing error detection studies suggest that not only are there valid and reliable methods for testing error detection ability, but certain variables have the potential to increase or decrease that ability. In addition, findings revealed that a tension exists between designing error detection studies with high ecological validity (real world, contextual relevance) and those with high internal validity (elimination of confounding variables). Based on these findings, I offer several recommendations for inservice music educators and music education faculty.

Prediction of 5-year Progression-Free Survival in Advanced Nasopharyngeal Carcinoma with Pretreatment PET/CT using Multi-Modality Deep Learning-based Radiomics

Purpose Deep Learning-based Radiomics (DLR) has achieved great success on medical image analysis. In this study, we aimed to explore the capability of our proposed end-to-end multi-modality DLR model using pretreatment PET/CT images to predict 5-year Progression-Free Survival (PFS) in advanced NPC.Methods A total of 170 patients with pathological confirmed advanced NPC (TNM stage III or IVa) were enrolled in this study. A 3D Convolutional Neural Network (CNN), with two branches to process PET and CT separately, was optimized to extract deep features from pretreatment multi-modality PET/CT images and use the derived features to predict the probability of 5-year PFS. Optionally, TNM stage, as a high-level clinical feature, can be integrated into our DLR model to further improve prognostic performance. Results For a comparison between Conventional Radiomic (CR) and DLR, 1456 handcrafted features were extracted, and three top CR methods, Random Forest (RF) + RF (AUC = 0.796 ± 0.009, testing error = 0.267 ± 0.007), RF + Adaptive Boosting (AdaBoost) (AUC = 0.783 ± 0.011, testing error = 0.286 ± 0.009), and L1-Logistic Regression (L1-LOG) + Kernel Support Vector Machines (KSVM) (AUC = 0.769 ± 0.008, testing error = 0.298 ± 0.006), were selected as benchmarks from 54 combinations of 6 feature selection methods and 9 classification methods. Compared to the three CR methods, our multi-modality DLR models using both PET and CT, with or without TNM stage (named PCT or PC model), resulted in the highest prognostic performance (PCT model: AUC = 0.842 ± 0.034, testing error = 0.194 ± 0.029; PC model: AUC = 0.825 ± 0.041, testing error = 0.223 ± 0.035). Furthermore, the multi-modality PCT model outperformed single-modality DLR models using only PET and TNM stage (named PT model: AUC = 0.818 ± 0.029, testing error = 0.218 ± 0.024) or only CT and TNM stage (named CT model: AUC = 0.657 ± 0.055, testing error = 0.375 ± 0.048). Conclusion Our study identified potential radiomics-based prognostic model for survival prediction in advanced NPC, and suggests that DLR could serve as a tool for aiding in cancer management.

Leveraging Neural Networks and Genetic Algorithms to Refine Electrode Properties in Redox Flow Batteries

Redox flow batteries are a nascent, yet promising, energy storage technology for which widespread deployment is hampered by technical and economic challenges. A performance-determining component in the reactor, present-day electrodes are often borrowed from adjacent electrochemical technologies rather than specifically designed for use in flow batteries. A lack of structural diversity in commercial offerings, coupled with the time constraints of wet-lab experiments, render broad electrode screening infeasible without a modeling complement. Herein, an experimentally validated model of a vanadium redox flow cell is used to generate polarization data for electrodes with different macrohomogeneous properties (thickness, porosity, volumetric surface area, and kinetic rate constant). Using these data sets, we then build and train a neural network with minimal average root-mean squared testing error (17.9 ± 1.8 mA cm⁻²) to compute individual parameter sweeps along the cell polarization curve. Finally, we employ a genetic algorithm with the neural network to ascertain electrode property values for improving cell power density. While the developed framework does not supplant experimentation, it is generalizable to different redox chemistries and may inform future electrode design strategies.

Leveraging Neural Networks and Genetic Algorithms to Refine Electrode Properties in Redox Flow Batteries

Redox flow batteries are a nascent, yet promising, energy storage technology for which widespread deployment is hampered by technical and economic challenges. A performance-determining component in the reactor, present-day electrodes are often borrowed from adjacent electrochemical technologies rather than specifically designed for use in flow batteries. A lack of structural diversity in commercial offerings, coupled with the time constraints of wet-lab experiments, render broad electrode screening infeasible without a modeling complement. Herein, an experimentally validated model of a vanadium redox flow cell is used to generate polarization data for electrodes with different macrohomogeneous properties (thickness, porosity, volumetric surface area, and kinetic rate constant). Using these data sets, we then build and train a neural network with minimal average root-mean squared testing error (17.9 ± 1.8 mA cm⁻²) to compute individual parameter sweeps along the cell polarization curve. Finally, we employ a genetic algorithm with the neural network to ascertain electrode property values for improving cell power density. While the developed framework does not supplant experimentation, it is generalizable to different redox chemistries and may inform future electrode design strategies.

Role of Capital Formation and Saving in Promoting Economic Growth in Nepal: An ARDL Bound Testing Approach

This study examines the relation and impact of gross capital formation and gross national saving on Nepal's Gross Domestic Product (GDP). It is based on the secondary data taken from various economics survey of Nepal and other published sources covering 33 data points from the fiscal year 1987/88 to 2019/20. Descriptive and empirical research designs examine the relation between GDP, gross capital formation, and gross national saving. The EViews 10 data processing software is used. Some econometrics tools like mean, dispersion, ARDL bound testing, error correction model, heteroskedasticity, serial correlation test, normality test, CUSUM test, and CUSUM square test are used. There is a long-run positive relationship between GDP and gross capital formation and gross national saving. The gross capital formation and gross national saving are individually and jointly significant to explain GDP in the long run, but there is a negative impact on regressors' GDP. Capital formation and saving positively impact GDP. Still, the effectiveness is not found satisfactory because a one percent increase in capital formation only increases GDP by 0.267 per cent. So, the saving amount must be utilized in the productive sector. The author of the research is not affected by the other researchers' findings, tools, and methods.