Search

This research paper delves into the challenges associated with deep learning models, specifically focusing on transfer learning. Despite the effectiveness of widely used models such as VGGNet, ResNet, and GoogLeNet, their deployment on resource-constrained devices is impeded by high memory bandwidth and computational costs, and to overcome these limitations, the study proposes pruning as a viable solution. Numerous parameters, particularly in fully connected layers, contribute minimally to computational costs, so we focus on convolution layers’ pruning. The research explores and evaluates three innovative pruning methods: the Max3 Saliency pruning method, the K-Means clustering algorithm, and the Singular Value Decomposition (SVD) approach. The Max3 Saliency pruning method introduces a slight variation by using the three maximum values of the kernel instead of all nine to compute the saliency score. This method is the most effective, substantially reducing parameter and Floating Point Operations (FLOPs) for both VGG16 and ResNet56 models. Notably, VGG16 demonstrates a remarkable 46.19% reduction in parameters and a 61.91% reduction in FLOPs. Using the Max3 Saliency pruning method, ResNet56 shows a 35.15% reduction in parameters and FLOPs. The K-Means pruning algorithm is also successful, resulting in a 40.00% reduction in parameters for VGG16 and a 49.20% reduction in FLOPs. In the case of ResNet56, the K-Means algorithm achieved a 31.01% reduction in both parameters and FLOPs. While the Singular Value Decomposition (SVD) approach provides a new set of values for condensed channels, its overall pruning ratio is smaller than the Max3 Saliency and K-Means methods. The SVD pruning method prunes 20.07% parameter reduction and a 24.64% reduction in FLOPs achieved for VGG16, along with a 16.94% reduction in both FLOPs and parameters for ResNet56. Compared with the state-of-the-art methods, the Max3 Saliency and K-Means pruning methods performed better in Flops reduction metrics.

Addressing real-time optimization problems becomes increasingly challenging as their complexity continues to escalate over time. So bio-optimization algorithms (BoAs) come into the picture to solve such problems due to their global search capability, adaptability, versatility, parallelism, and robustness. This article aims to perform hyperparameter tuning of machine learning (ML) models by integrating them with BoAs. Aiming to maximize the accuracy of the hybrid bio-optimized defect prediction (HBoDP) model, this research paper develops four novel hybrid BoAs named the gravitational force Lévy flight grasshopper optimization algorithm (GFLFGOA), the gravitational force Lévy flight grasshopper optimization algorithm–sparrow search algorithm (GFLFGOA-SSA), the gravitational force grasshopper optimization algorithm–sparrow search algorithm (GFGOA-SSA), and the Lévy flight grasshopper optimization algorithm–sparrow search algorithm (LFGOA-SSA). These aforementioned algorithms are proposed by integrating the good exploration capacity of the SSA with the faster convergence of the LFGOA and GFGOA. The performances of the GFLFGOA, GFLFGOA-SSA, GFGOA-SSA, and LFGOA-SSA are verified by conducting two different experiments. Firstly, the experimentation was conducted on nine benchmark functions (BFs) to assess the mean, standard deviation (SD), and convergence rate. The second experiment focuses on boosting the accuracy of the HBoDP model through the fine-tuning of the hyperparameters in the artificial neural network (ANN) and XGBOOST (XGB) models. To justify the effectiveness and performance of these hybrid novel algorithms, we compared them with four base algorithms, namely the grasshopper optimization algorithm (GOA), the sparrow search algorithm (SSA), the gravitational force grasshopper optimization algorithm (GFGOA), and the Lévy flight grasshopper optimization algorithm (LFGOA). Our findings illuminate the effectiveness of this hybrid approach in enhancing the convergence rate and accuracy. The experimental results show a faster convergence rate for BFs and improvements in software defect prediction accuracy for the NASA defect datasets by comparing them with some baseline methods.

The rapidly increasing scientific research on the stock market and the visible impact of media on equity prices are nowadays in limelight. To a greater extent, causal analysis can reckon the sentimental effect of the broadcasted content on stock valuation. We propose a four stage model to detect the direction of information flow between the news sentiment and stock price. Whilst web scraping explores and extracts the news datasets, the modified VADER algorithm finds the sentiments of the aired media. The associational causal analysis determines the cause effect between the news and stock price. The results suggest that the non parametric Shannon and Renyi's entropy approach supersedes the Granger test, a parametric study which is constrained to Gaussian time series with linear causation. Since Renyi's Entropy can perfectly identify the deluge of information during quick leaps, it is regarded as a beneficial formulation for investors when evaluating stocks with a fewer number of news mentions. The impact of news during the COVID-19 pandemic over the pharmaceutical sector was also done. The study infers an explicit information flow and direction of causality between news sentiment and stock price movement, which can be used to devise future investment and consumption strategies.

Addressing real-time optimization problems becomes increasingly challenging as their complexity continues to escalate over time. So bio-optimization algorithms (BoAs) come into the picture to solve such problems due to their global search capability, adaptability, versatility, parallelism, and robustness. This article aims to perform hyperparameter tuning of machine learning (ML) models by integrating them with BoAs. Aiming to maximize the accuracy of the hybrid bio-optimized defect prediction (HBoDP) model, this research paper develops four novel hybrid BoAs named the gravitational force Lévy flight grasshopper optimization algorithm (GFLFGOA), the gravitational force Lévy flight grasshopper optimization algorithm–sparrow search algorithm (GFLFGOA-SSA), the gravitational force grasshopper optimization algorithm–sparrow search algorithm (GFGOA-SSA), and the Lévy flight grasshopper optimization algorithm–sparrow search algorithm (LFGOA-SSA). These aforementioned algorithms are proposed by integrating the good exploration capacity of the SSA with the faster convergence of the LFGOA and GFGOA. The performances of the GFLFGOA, GFLFGOA-SSA, GFGOA-SSA, and LFGOA-SSA are verified by conducting two different experiments. Firstly, the experimentation was conducted on nine benchmark functions (BFs) to assess the mean, standard deviation (SD), and convergence rate. The second experiment focuses on boosting the accuracy of the HBoDP model through the fine-tuning of the hyperparameters in the artificial neural network (ANN) and XGBOOST (XGB) models. To justify the effectiveness and performance of these hybrid novel algorithms, we compared them with four base algorithms, namely the grasshopper optimization algorithm (GOA), the sparrow search algorithm (SSA), the gravitational force grasshopper optimization algorithm (GFGOA), and the Lévy flight grasshopper optimization algorithm (LFGOA). Our findings illuminate the effectiveness of this hybrid approach in enhancing the convergence rate and accuracy. The experimental results show a faster convergence rate for BFs and improvements in software defect prediction accuracy for the NASA defect datasets by comparing them with some baseline methods. [ABSTRACT FROM AUTHOR]

abstract: Polymer matrix nanocomposites are in high demand and have attracted great interest in the expansion of new functional materials owing to their potential in environmental remediation applications. In this context, Polypyrrole-Based Copper oxide-Nickel oxide Nanocomposite (PPCN) was successfully fabricated by combining chemical co-precipitation and in-situ polymerization methods. The chemical co-precipitation technique was used to synthesize Copper oxide-Nickel oxide Nanocomposite using citric acid as a capping agent and in-situ polymerization method was applied to produce Polypyrrole (PPY). The photocatalytic process is a green technology that is developing as a viable option, for removing a variety of toxic pollutants, and antibacterial activity in surface water irrigation is beneficial for the protection of the edible crops from the microbial and dye pollutants in wastewater. The synthesized PPCN is investigated for its photocatalytic properties and for antibacterial activities and is safe for the environment making them a promising material for wide-ranging applications including wastewater treatment, bacterial control, and agricultural purposes. The photocatalytic behavior of the PPCN was studied for the removal of dyes under UV irradiation and showed that these polymer-based nanocomposites could remove and destroy the dye contaminants in high worth. Moreover, the PPCN acts as an excellent catalyst and observed a remarkable dye degradation efficiency of 86% in Alizarin Red (AR) and 80% in Toluidine Blue (TB) dyes within 5hrs. The results show that the PPCN could be used as a well-performing catalyst for water contamination removal. In addition, the antibacterial results indicate that the maximum antibacterial action is against Staphylococcus aureus (S. aureus), Bacillus cereus (B. cereus), and Escherichia coli (E. coli).

Over the past decade, there has been renewed interest in and commitment to resolving the endemic problem of statelessness, most clearly exemplified by the United Nations High Commissioner for Refugees’ Global Action Plan to End Statelessness 2014-24, which sets out to end statelessness by 2024. Despite the plethora of recent attention to questions of citizenship, its converse, the problem of statelessness and its effect on children, has not been adequately investigated. This paper attempts to delineate the causes of childhood statelessness in particular and to analyze the international legal framework for reducing and preventing it. It examines how statelessness is created, how it persists and why it brings with it the deprivations it does. It then subjects the customary and modern international legal norms governing childhood statelessness and enforcement strategies at Global level to close scrutiny and identifies the clearly discernible drawbacks and road blocks. It concludes with suggestions, inter alia, to make the jus soli citizenship a mandatory default clause in the citizenship laws of every country, to further prioritize birth registration and data collection and to strengthen the UPR process and reporting procedure.