Differentially private SGD with random features

In the realm of large-scale machine learning,it is crucial to explore methods for reducing computational complexity and memory demands while maintaining generalization performance.Additionally,since the collected data may contain some sensitive information,it is also of great significance to study privacy-preserving machine learning algorithms.This paper focuses on the performance of the differentially private stochastic gradient descent(SGD)algorithm based on random features.To begin,the algorithm maps the original data into a lowdimensional space,thereby avoiding the traditional kernel method for large-scale data storage requirement.Subsequently,the algorithm iteratively optimizes parameters using the stochastic gradient descent approach.Lastly,the output perturbation mechanism is employed to introduce random noise,ensuring algorithmic privacy.We prove that the proposed algorithm satisfies the differential privacy while achieving fast convergence rates under some mild conditions.

Bioactive compounds in Hericium erinaceus and their biological properties:a review

Hericium erinaceus is a nutritious edible and medicinal fungi,rich in a variety of functional active ingredients,with various physiological functions such as antioxidation,anticancer,and enhancing immunity.It is also effective in protecting the digestive system and preventing neurodegenerative diseases.In this review paper,we summarize the sources,structures and efficacies of the main active components in H.erinaceus fruiting body,mycelium,and culture media,and update the latest research progress on their biological activities and the related molecular mechanisms.Based on this information,we provide detailed challenges in current research,industrialization and information on the active ingredients of H.erinaceus.Perspectives for future studies and new applications of H.erinaceus are proposed.

Soil bacterial and fungal communities resilience to long-term nitrogen addition in subtropical forests in China

Atmospheric nitrogen(N)deposition is predicted to increase,especially in the subtropics.However,the responses of soil microorganisms to long-term N addition at the molecular level in N-rich subtropical forests have not been clarified.A long-term nutrient addition experiment was conducted in a subtropical evergreen old-growth forest in China.The four treatments were:control,low N(50 kg N ha^(-1)a^(-1)),high N(100 kg N ha^(-1)a^(-1)),and combined N and phosphorus(P)(100 kg N ha^(-1)a^(-1)+50 kg P ha^(-1)a^(-1)).Metagenomic sequencing characterized diversity and composition of soil microbial communities and used to construct bacterial/fungal co-occurrence networks.Nutrient-treated soils were more acidic and had higher levels of dissolved organic carbon than controls.There were no significant differences in microbial diversity and community composition across treatments.The addition of nutrients increased the abundance of copiotrophic bacteria and potentially beneficial microorganisms(e.g.,Gemmatimonadetes,Chaetomium,and Aureobasidium).Low N addition increased microbiome network connectivity.Three rare fungi were identified as module hubs under nutrient addition,indicating that low abundance fungi were more sensitive to increased nutrients.The results indicate that the overall composition of microbial communities was stable but not static to long-term N addition.Our findings provide new insights that can aid predictions of the response of soil microbial communities to long-term N addition.

Fractional Gradient Descent RBFNN for Active Fault-Tolerant Control of Plant Protection UAVs

With the increasing prevalence of high-order systems in engineering applications, these systems often exhibitsignificant disturbances and can be challenging to model accurately. As a result, the active disturbance rejectioncontroller (ADRC) has been widely applied in various fields. However, in controlling plant protection unmannedaerial vehicles (UAVs), which are typically large and subject to significant disturbances, load disturbances andthe possibility of multiple actuator faults during pesticide spraying pose significant challenges. To address theseissues, this paper proposes a novel fault-tolerant control method that combines a radial basis function neuralnetwork (RBFNN) with a second-order ADRC and leverages a fractional gradient descent (FGD) algorithm.We integrate the plant protection UAV model’s uncertain parameters, load disturbance parameters, and actuatorfault parameters and utilize the RBFNN for system parameter identification. The resulting ADRC exhibits loaddisturbance suppression and fault tolerance capabilities, and our proposed active fault-tolerant control law hasLyapunov stability implications. Experimental results obtained using a multi-rotor fault-tolerant test platformdemonstrate that the proposed method outperforms other control strategies regarding load disturbance suppressionand fault-tolerant performance.

Smart prediction of liquefaction-induced lateral spreading

The prediction of liquefaction-induced lateral spreading/displacement(Dh)is a challenging task for civil/geotechnical engineers.In this study,a new approach is proposed to predict Dh using gene expression programming(GEP).Based on statistical reasoning,individual models were developed for two topographies:free-face and gently sloping ground.Along with a comparison with conventional approaches for predicting the Dh,four additional regression-based soft computing models,i.e.Gaussian process regression(GPR),relevance vector machine(RVM),sequential minimal optimization regression(SMOR),and M5-tree,were developed and compared with the GEP model.The results indicate that the GEP models predict Dh with less bias,as evidenced by the root mean square error(RMSE)and mean absolute error(MAE)for training(i.e.1.092 and 0.815;and 0.643 and 0.526)and for testing(i.e.0.89 and 0.705;and 0.773 and 0.573)in free-face and gently sloping ground topographies,respectively.The overall performance for the free-face topology was ranked as follows:GEP>RVM>M5-tree>GPR>SMOR,with a total score of 40,32,24,15,and 10,respectively.For the gently sloping condition,the performance was ranked as follows:GEP>RVM>GPR>M5-tree>SMOR with a total score of 40,32,21,19,and 8,respectively.Finally,the results of the sensitivity analysis showed that for both free-face and gently sloping ground,the liquefiable layer thickness(T_(15))was the major parameter with percentage deterioration(%D)value of 99.15 and 90.72,respectively.

Biology of Flowering of Some Species of the Genus Phlomoides Moench

The article presents the biology of flowering and the daily dynamics of flowering of two species from the Lamiaceae family: ph. anisochila va ph. sogdiana of the distribution out in the Nuratau Mountains range.

Characterization of Wastewater in School Environments for an Ecological Treatment Solution: A Case Study of Ndiebene Gandiol 1 School

The study conducted at Ndiebene Gandiol 1 school in Senegal has unveiled serious environmental and public health challenges. The wastewater analysis revealed high levels of Biochemical Oxygen Demand (BOD5), Chemical Oxygen Demand (COD), and fecal coliforms, signaling potential risks to the well-being of students and staff. This situation mirrors a wider issue in rural educational settings, where inadequate sanitation persists. Intensive wastewater treatment options are known for their effectiveness against high pollutant loads but are resource-intensive in both energy and cost. Conversely, extensive treatment systems, while requiring more land, provide a sustainable alternative by harnessing natural processes for pollutant removal. The research suggests a hybrid treatment approach could serve the school’s needs, balancing the robust capabilities of intensive methods with the ecological benefits of extensive systems. Such a solution would need to be tailored to the specific environmental, financial, and logistical context of the school, based on comprehensive feasibility studies and stakeholder engagement. This study’s findings underscore the urgency of addressing sanitation in schools, as it is intrinsically linked to the health and academic success of students. Quick, effective, and long-term strategies are vital to secure a healthier and more prosperous future for the youth. With proper implementation, the school can transform its sanitation facilities, setting a precedent for rural educational institutions in Senegal and similar contexts globally.

Design and Sizing of an Ecological Wastewater Treatment System in a School Environment: A Case Study of Ndiebene Gandiol 1 School

The primary objective of this study was to design and size a sustainable sanitation solution for the Ndiebene Gandiol 1 school located in the eponymous commune in northern Senegal. Field investigations led to the collection of wastewater samples. Their analysis revealed specific pollutant loads, including loads of BOD5 3.6966 kgO2/day and COD of 12.8775 kgO2/day, which were central to the design phase. Following a rigorous assessment of the existing sanitation infrastructure, constructed wetland (CWs) emerged as the most appropriate ecological solution. This system, valued for its ability to effectively remove contaminants, was tailored to the specific needs of the site. Consequently, the final design of the filter extends over 217.16 m2, divided into two cells of 108.58 m2 each, with dimensions of 12.77 m in length and 8.5 m in width. The depth of the filtering medium is approximately 0.60 m, meeting the standards while ensuring maximized purification. Typha, an indigenous and prolific plant known for its purification abilities, was selected as the filtering agent. Concurrently, non-crushed gravel was chosen for its proven filtration capacity. This study is the result of a combination of scientific rigor and design expertise. It provides a holistic view of sanitation for Ndiebene Gandiol. The technical specifications and dimensions of the constructed wetland filter embody an approach that marries indepth analysis and practical application, all aimed at delivering an effective and long-lasting solution to the local sanitation challenges. By integrating precise scientific data with sanitation design expertise, this study delivers a holistic solution for Ndiebene Gandiol. The detailed dimensions and specifications of the constructed wetland filter reflect a methodology that combines meticulous analysis with practical adaptation, aiming to provide an effective and sustainable response to the challenges of rural and school sanitation in the northern region of Senegal.

Spark Plasma Sintering of Boron Carbide Using Ti_(3)SiC_(2) as a Sintering Additive

Boron carbide has unique properties for wide application possibilities;however,poor sinterability limits its applications.One approach to overcome this limitation is the addition of secondary phases into boron carbide.Boron carbide based composite ceramics are produced by the direct addition of secondary phases into the structure or via reactive sintering using a sintering additive.The present study investigated the effect of Ti_(3)SiC_(2) addition to boron carbide by reactive spark plasma sintering in the range of 1700-1900℃.Ti_(3)SiC_(2) phase decomposed at high temperatures and reacted with B4C to form secondary phases of TiB2 and SiC.The results demonstrated that the increase of Ti_(3)SiC_(2) addition(up to 15 vol%)effectively promoted the densification of B4C and yielded higher hardness.However,as the amount of Ti_(3)SiC_(2) increased further,the formation of microstructural inhomogeneity and agglomeration of secondary phases caused a decrease in hardness.

Ecological Wastewater Treatment System Using a Horizontal Flow Biological Reactor: The Case of Vetiver

Confronted with the challenge of wastewater management, particularly in the school environment of Senegal, our study set out to achieve multiple objectives. Following field surveys, laboratory analyses of wastewater samples were carried out, revealing a significant pollutant load. In the community of Gandiol, near Saint-Louis (Senegal), the school of Ndiebene Gandiol 1 faces significant sanitation challenges. Our study aimed to address this issue by using a constructed filter composed of two filtering bed cells measuring 12 × 8.5 m, preceded by a septic tank. We particularly focused on the influence of Vetiver;a plant chosen for its purification potential. Our analyses showed remarkable efficiency of the filter. Elimination rates reached 95% for 5-Day Biochemical Oxygen Demand (BOD5), 91% for Chemical Oxygen Demand (COD), and 92% for SS, far exceeding the Senegalese standards set at 50 mg/L, 200 mg/L, and 40 mg/L, respectively. Furthermore, the concentration of fecal coliforms was reduced to 176 FCU/100mL, well below the Senegalese threshold of 2000 FCU/100mL and close to the World Health Organization’s (WHO) recommendation of 1000 FCU/100mL. However, despite these promising results, some parameters, particularly the concentration of certain pollutants, approached the thresholds defined by European legislation. For example, for Suspended Solids (SS), the post-treatment level of 3 mg/L was well below the Senegalese standard but edged close to the European minimum of 10 mg/L. In conclusion, the Vetiver filter demonstrated a remarkable ability to treat school wastewater, offering high pollutant elimination percentages. These results suggest significant opportunities for the reuse of treated water, potentially in areas such as irrigation, though some adjustments may be necessary to meet the strictest standards such as those of the European union (EU).

Near-Field Beam Training for Holographic MIMO Communications: Typical Methods, Challenges and Future Directions

Holographic multiple-input multiple-output(HMIMO)has become an emerging technology for achieving ultra-high frequency spectral efficiency and spatial resolution in future wireless systems.The increasing antenna aperture leads to a more significant characterization of the spherical wavefront in near-field communications in HMIMO scenarios.Beam training as a key technique for wireless communication is worth exploring in this near-field scenario.Compared with the widely researched far-field beam training,the increased dimensionality of the search space for near-field beam training poses a challenge to the complexity and accuracy of the proposed algorithm.In this paper,we introduce several typical near-field beam training methods:exhaustive beam training,hierarchical beam training,and multi-beam training that includes equal interval multi-beam training and hash multi-beam training.The performances of these methods are compared through simulation analysis,and their effectiveness is verified on the hardware testbed as well.Additionally,we provide application scenarios,research challenges,and potential future research directions for near-field beam training.

Performance Evaluation of Two Series Vertical Flow Filters for Wastewater Treatment: A Case Study of the Prototype Installed at Gaston Berger University, Saint-Louis, Senegal

This paper evaluates the efficacy of two sequential vertical flow filters (VFF), FV1 and FV2, implanted with Typha, in a pilot-scale wastewater treatment system. FV1 comprises three cells (FV1a, FV1b, and FV1c), while FV2 consists of two cells (FV2a and FV2b), each designed to reduce various physicochemical and microbiological pollutants from wastewater. Quantitative analyses show significant reductions in electrical conductivity (from 1331 to 1061 μS/cm), biochemical oxygen demand (BOD5 from 655.6 to 2.3 mg/L), chemical oxygen demand (COD from 1240 to 82.2 mg/L), total nitrogen (from 188 to 37.3 mg/L), and phosphates (from 70.9 to 14.6 mg/L). Notably, FV2 outperforms FV1, particularly in decreasing dissolved salts and BOD5 to remarkably low levels. Microbiological assessments reveal a substantial reduction in fecal coliforms, from an initial concentration of 7.5 log CFU/100mL to 3.7 log CFU/100mL, and a complete elimination of helminth eggs, achieving a 100% reduction rate in FV2. The study highlights the impact of design parameters, such as filter material, media depth, and plant species selection, on treatment outcomes. The findings suggest that the judicious choice of these components is critical for optimizing pollutant removal. For instance, different filtration materials show varying efficacies, with silex plus river gravel in FV1c achieving superior pollutant reduction rates. In conclusion, VFFs emerge as a promising solution for wastewater treatment, underscoring the importance of design optimization to enhance system efficiency. Continuous monitoring and adaptation of treatment practices are imperative to ensure water quality, allowing for safe environmental discharge or water reuse. The research advocates for ongoing improvements in wastewater treatment technologies, considering the environmental challenges of the current era. The study concludes with a call for further research to maximize the effectiveness of VFFs in water management.

A Deep Reinforcement Learning-Based Technique for Optimal Power Allocation in Multiple Access Communications

Formany years,researchers have explored power allocation(PA)algorithms driven bymodels in wireless networks where multiple-user communications with interference are present.Nowadays,data-driven machine learning methods have become quite popular in analyzing wireless communication systems,which among them deep reinforcement learning(DRL)has a significant role in solving optimization issues under certain constraints.To this purpose,in this paper,we investigate the PA problem in a k-user multiple access channels(MAC),where k transmitters(e.g.,mobile users)aim to send an independent message to a common receiver(e.g.,base station)through wireless channels.To this end,we first train the deep Q network(DQN)with a deep Q learning(DQL)algorithm over the simulation environment,utilizing offline learning.Then,the DQN will be used with the real data in the online training method for the PA issue by maximizing the sumrate subjected to the source power.Finally,the simulation results indicate that our proposedDQNmethod provides better performance in terms of the sumrate compared with the available DQL training approaches such as fractional programming(FP)and weighted minimum mean squared error(WMMSE).Additionally,by considering different user densities,we show that our proposed DQN outperforms benchmark algorithms,thereby,a good generalization ability is verified over wireless multi-user communication systems.

Photoelectric State with Long-Term Relaxation in CdTe:(Ag, Cu, Cd) and Sb2Se3:Se Photovoltaic Films

The results of an experimental study of long-term relaxation of the photoelectret state of polycrystalline CdTe:(Ag, Cu, Cd) and Sb2Se3:Se films with an anomalous photovoltaic property are presented. In such films, the residual photovoltage is caused by the separation of photocarriers by the built-in electrostatic field of the near-surface region of space charges and their asymmetric capture by deep levels of impurities or complexes, including impurity atoms and intrinsic defects, both in the bulk and on the surface of crystal grains. It has been shown that in activated films, a two-step exponential temporary relaxation of the initial photovoltage of the order of VAPV ≈ (500-600) V is detected, and only 10% of it experiences long-term relaxation (t ≈ 100-120 min).

Optimization of Biofuel Formulation by Mixture Design

With the full growth of energy needs in the world, several studies are now focused on finding renewable sources. The aim of this work is to optimise biofuel formulation from a mixture design by studying physical properties, such as specific gravity and kinematic viscosity of various formulated mixtures. Optimization from the mixture plan revealed that in the chosen experimental domain, the optimal conditions are: 40% for used frying oil (UFO), 50% for bioethanol and 10% for diesel. These experimental conditions lead to a biofuel with a density of 0.84 and a kinematic viscosity of 2.97 cSt. These parameters are compliant with the diesel quality certificate in tropical areas. These density and viscosity values were determined according to respective desirability values of 0.68 and 0.75.

Least Square Finite Element Model for Analysis of Multilayered Composite Plates under Arbitrary Boundary Conditions

Laminated composites are widely used in many engineering industries such as aircraft, spacecraft, boat hulls, racing car bodies, and storage tanks. We analyze the 3D deformations of a multilayered, linear elastic, anisotropic rectangular plate subjected to arbitrary boundary conditions on one edge and simply supported on other edge. The rectangular laminate consists of anisotropic and homogeneous laminae of arbitrary thicknesses. This study presents the elastic analysis of laminated composite plates subjected to sinusoidal mechanical loading under arbitrary boundary conditions. Least square finite element solutions for displacements and stresses are investigated using a mathematical model, called a state-space model, which allows us to simultaneously solve for these field variables in the composite structure’s domain and ensure that continuity conditions are satisfied at layer interfaces. The governing equations are derived from this model using a numerical technique called the least-squares finite element method (LSFEM). These LSFEMs seek to minimize the squares of the governing equations and the associated side conditions residuals over the computational domain. The model is comprised of layerwise variables such as displacements, out-of-plane stresses, and in- plane strains, treated as independent variables. Numerical results are presented to demonstrate the response of the laminated composite plates under various arbitrary boundary conditions using LSFEM and compared with the 3D elasticity solution available in the literature.

Artificial Intelligence-Enhanced Learning:A New Paradigm in the“Business Data Analysis and Application”Course

This paper explores the transformative impact of generative artificial intelligence(AI)on the“Business Data Analysis and Application”course in the post-2023 era,marking a significant paradigm shift in educational methodologies.It investigates how generative AI reshapes teaching and learning dynamics,enhancing the processing of complex data sets and nurturing critical thinking skills.The study highlights the role of AI in fostering dynamic,personalized,and adaptive learning experiences,addressing the evolving pedagogical needs of the business sector.Key challenges,including equitable access,academic integrity,and ethical considerations such as data privacy and algorithmic bias,are thoroughly examined.The research reveals that the integration of generative AI aligns with current professional demands,equipping students with cutting-edge AI tools,and tailoring learning to individual needs through real-time feedback mechanisms.The study concludes that the incorporation of generative AI into this course signifies a substantial evolution in educational approaches,offering profound implications for student learning and professional development.

Non-Linear Matrix Completion

Current methods for predicting missing values in datasets often rely on simplistic approaches such as taking median value of attributes, limiting their applicability. Real-world observations can be diverse, taking stock price as example, ranging from prices post-IPO to values before a company’s collapse, or instances where certain data points are missing due to stock suspension. In this paper, we propose a novel approach using Nonlinear Matrix Completion (NIMC) and Deep Matrix Completion (DIMC) to predict associations, and conduct experiment on financial data between dates and stocks. Our method leverages various types of stock observations to capture latent factors explaining the observed date-stock associations. Notably, our approach is nonlinear, making it suitable for datasets with nonlinear structures, such as the Russell 3000. Unlike traditional methods that may suffer from information loss, NIMC and DIMC maintain nearly complete information, especially in high-dimensional parameters. We compared our approach with state-of-the-art linear methods, including Inductive Matrix Completion, Nonlinear Inductive Matrix Completion, and Deep Inductive Matrix Completion. Our findings show that the nonlinear matrix completion method is particularly effective for handling nonlinear structured data, as exemplified by the Russell 3000. Additionally, we validate the information loss of the three methods across different dimensionalities.

A Comparative Analysis of Vetiver and Typha in Ecological Wastewater Treatment Using a Horizontal Flow Constructed Wetlands in Rural Setting

This study presents an assessment of wastewater ecological treatment processes utilizing a horizontal flow bio-reactor at the Ndiebene Gandiol 1 school. It primarily aims to juxtapose the filtration efficacy of two distinct vegetative cells, Vetiver and Typha, in the pursuit of sustainable wastewater management strategies for rural scholastic institutions. A synergistic approach was employed, integrating on-site surveys for site-specific insights and laboratory analyses to quantify the pollutant loads pre- and post-treatment. Our findings indicate that both Vetiver and Typha-infused filter beds significantly reduce most contaminants, with particular success in diminishing chemical oxygen demand (COD) and biological oxygen demand (BOD5). Vetiver was notable for its superior reduction of COD, achieving an average effluent concentration of 74 mg/L, in contrast to Typha’s 155 mg/L. Conversely, Typha excelled in suspended solids removal, registering 1 mg/L against Vetiver’s 3 mg/L. While both systems notably surpassed the target metrics across several indicators, including fecal coliform reduction, our results pinpoint the need for refinement in phosphate remediation. Conclusively, the study underscores the efficacy of both Vetiver and Typha systems in rural wastewater treatment contexts, with their integrative application potentially paving the way for enhanced system robustness and efficiency. The outcomes herein highlight the imperative for continued research to further hone these ecological treatment modalities, especially concerning phosphate elimination.

An Approach to Detect Structural Development Defects in Object-Oriented Programs

Structural development defects essentially refer to code structure that violates object-oriented design principles. They make program maintenance challenging and deteriorate software quality over time. Various detection approaches, ranging from traditional heuristic algorithms to machine learning methods, are used to identify these defects. Ensemble learning methods have strengthened the detection of these defects. However, existing approaches do not simultaneously exploit the capabilities of extracting relevant features from pre-trained models and the performance of neural networks for the classification task. Therefore, our goal has been to design a model that combines a pre-trained model to extract relevant features from code excerpts through transfer learning and a bagging method with a base estimator, a dense neural network, for defect classification. To achieve this, we composed multiple samples of the same size with replacements from the imbalanced dataset MLCQ1. For all the samples, we used the CodeT5-small variant to extract features and trained a bagging method with the neural network Roberta Classification Head to classify defects based on these features. We then compared this model to RandomForest, one of the ensemble methods that yields good results. Our experiments showed that the number of base estimators to use for bagging depends on the defect to be detected. Next, we observed that it was not necessary to use a data balancing technique with our model when the imbalance rate was 23%. Finally, for blob detection, RandomForest had a median MCC value of 0.36 compared to 0.12 for our method. However, our method was predominant in Long Method detection with a median MCC value of 0.53 compared to 0.42 for RandomForest. These results suggest that the performance of ensemble methods in detecting structural development defects is dependent on specific defects.