Explore the vast range of titles available.

In intrusion detection, imbalanced datasets often degrade the accuracy of certain detections, potentially allowing malicious traffic to remain undetected and leading to significant losses. The multi-layer perceptron (MLP) offers distinct advantages for intrusion detection, as attack patterns often follow complex, nonlinear relationships. These patterns can be effectively captured through MLP’s multiple nonlinear transformations, such as ReLU and Sigmoid activation functions, which are especially beneficial for intrusion detection. Additionally, MLP exhibits low resource consumption, making it suitable for resource-constrained environments. However, MLP often struggles to accurately classify minority classes in imbalanced datasets due to its limited feature extraction capabilities. In contrast, convolutional neural networks (CNNs), particularly AlexNet’s small convolutional filters, offer more precise feature extraction for detailed dataset features. Therefore, this study integrates AlexNet’s feature extraction module with MLP and incorporates the SKNet attention mechanism to improve the recognition of minority classes. Experimental results show that our enhanced MLP algorithm outperforms the standard MLP across all seven proposed classification tasks. Specifically, the F1 scores for BotnetARES and PortScan show significant improvements of 18.93% and 26.57%, respectively, validating the efficacy of the algorithmic enhancements.

Based on the characteristics of underdeveloped areas, this paper selects the panel data of 15 underdeveloped counties in Anhui Province from 2013 to 2019 and uses the panel threshold model to empirically analyze the sustainability of rural tourism development. The results show that: (1) Rural tourism development has a non-linear positive impact on poverty alleviation in underdeveloped areas and has a double threshold effect. (2) When the poverty rate is used to express the poverty level, it can be found that the development of rural tourism at a high level can significantly promote poverty alleviation. (3) When the number of poor people is used to express the poverty level, it can be found that the poverty reduction effect shows a marginal decreasing trend with the phased improvement of the development level of rural tourism. (4) The degree of government intervention, industrial structure, economic development, and fixed asset investment play a more significant role in poverty alleviation. Therefore, we believe that we need to actively promote rural tourism in underdeveloped areas, establish a mechanism for the distribution and sharing of rural tourism benefits, and form a long-term mechanism for rural tourism poverty reduction.

In this paper, a double-layer patterned graphene-based frequency-selective absorber (DGFSA) is proposed as a means of reducing an antenna’s radar cross-section (RCS) while simultaneously increasing its gain. The antenna consists of a patch antenna with Multi-Graphene Frequency Selective Absorber (MGFSA) mounted on top. The DGFSA consists of double-layer patterned graphene and a band-pass frequency selective surface (FSS). Two patterned graphene lossy layers with different square resistances are used, which broaden the electromagnetic (EM) wave absorption bandwidth of the DGFSA, thus greatly reducing the out-band monostatic RCSs of the patch antenna. Meanwhile, due to the quasi-Fabry-Perot (F-P) effect, the gain of the proposed antenna is enhanced by 2.4 dB. Additionally, the low-RCS antenna reduces the monostatic RCS from 1.32 to 17 GHz under y-polarization and from 1.4 to 16.8 GHz under x-polarization, respectively. Furthermore, a decrease in the bistatic RCS is accomplished. Results from simulations and measurements match up nicely, which means the antenna we proposed has a good application on the stealth platform.

Fertilizer nitrogen (N) application has been shown to impact methane (CH 4 ) emissions, yield and quality from rice cropping systems, yet the responses of CH 4 fluxes, yield and quality to N reduction and combined application of organic fertilizer in subtropical rice cropping systems are not well documented. Six experimental treatments were conducted: N90 kg N ha -1 of urea (N1), organic fertilizer with equal N90 (O1) and 80% urea + 20% organic fertilizer (N1O1), farmer’s common practice with N270 kg N ha -1 of urea (N2), organic fertilizer with equal N270 (O2) and 80% urea + 20% organic fertilizer (N2O2) were conducted to simultaneously measure the CH 4 flux, yield and quality from a subtropical rice cropping system in south China. Results showed that increased N fertilizer application significantly stimulated soil CH 4 emission, increased rice yield and altered quality in paddy fields. CH 4 emissions were quantified under different N fertilizer management practices in the peak rice growing season during the tillering and heading stages, respectively. Organic fertilizer alone increased CH 4 emission by 442.1% in O1 and by 337.3% in O2 compared with urea. However, relative to organic fertilizer, organic fertilizer combined with urea significantly decreased CH 4 emissions by 48.4% in O1 and by 39.2% in O2. Compared with N1 and N2 treatment, rice yield was significantly decreased by 34.4% and 39.5% under O1 and O2, while significantly enhanced by 49.8% and 22.3%, respectively, under N1O1 and N2O2 ( P < 0.05). The protein content significantly increased under N1O1 by 18.8% and 41.5%, the amylose content by 30.3% and 14.8%, and the gel consistency by 32.7% and 15.5% in contrast to N1 and O1 ( P < 0.05). Similarly, the protein content, amylose content and gel consistency under N2O2 were consistent with the rice quality under the N1O1 treatments above. In summary, optimizing organic fertilizer combined with urea practices was a win-win strategy to improve grain yield and quality while reducing CH 4 emissions in the rice cropping system. This study provides new insights into the fertilizer types on CH4 emission and rice production of rice cropping systems.

With the advent of the Internet, movie reviews have emerged as a crucial reference for users in selecting films and hold significant value in guiding filmmakers and platforms in content recommendation. Consequently, accurate classification of movie reviews has extensive practical applications. Traditional manual classification methods, however, are not only time-intensive and laborious but also susceptible to subjective bias. In response, automated classification techniques leveraging deep learning have become a promising alternative. Among these, the BERT model, renowned for its bidirectional encoder architecture, excels in contextual understanding and semantic representation. Nevertheless, it faces challenges in capturing long-range word dependencies and fully extracting local features in lengthy texts. Moreover, model bias stemming from sensitive information, such as gender and race, embedded in the data can compromise the fairness of classification outcomes. To address these limitations, this study introduces several enhancements to the BERT model. First, a dynamic positional offset encoding mechanism grounded in attention is employed to replace traditional absolute positional encoding, thereby enhancing the model’s capacity to process positional information. Second, a dynamic weighted fusion pooling strategy is proposed, integrating average pooling, maximum pooling, and self-attention pooling to improve the comprehensiveness of feature extraction. Additionally, during data preprocessing, sensitive attributes such as gender and race are mitigated through the removal or obfuscation of specific terms or features, combined with data augmentation techniques including easy data augmentation (EDA) and noise injection to generate neutral review samples. This approach reduces potential biases and enhances the model’s generalization capabilities. Experimental results on the IMDb movie review dataset demonstrate the efficacy of the proposed improvements, with the improved BERT model achieving a 0.73% increase in F1 score and a 0.90% improvement in accuracy, thereby validating the effectiveness of the modifications.

Passive temperature sensing systems based on the Internet of Things (IoT) present an efficient, reliable, and convenient solution for temperature monitoring with extensive application prospects and market value. This paper introduces a passive, battery-free, chipless, metasurface temperature sensing tag. The key insight is that the sensing tag uses vanadium dioxide ( ) to solve the problems of measuring distance, large size, and high cost related to active devices. The sensing tag fabricated with tungsten-doped powder demonstrated a significant variation in the reflection magnitude within the temperature range of 34–42 °C. It was achieved through coating, sintering, metasurface design, and ion beam etching. Experimental results showed that the square resistance of the prepared coating decreased from 1003 to 90 as the temperature increased from 34 to 42 °C. Additionally, the reflection magnitude of the tag significantly increased with the temperature decrease in the 3.5–5.27 GHz frequency band. These results indicate that the passive temperature sensing tags can achieve rapid and accurate temperature sensing within the 34–42 °C range.

Fused deposition modeling (FDM) technology is widely used in the areas of rapid prototyping, education, automobile and health care. However, the disadvantages of high porosity, low surface quality, and significant anisotropic mechanical properties hinder the applications in high-demand fields. In this paper, numerical model and experimental tests are combined to study the effects of layer thickness and extrusion ratio on the printing quality and mechanical properties of FDM structures. The computational fluid dynamics (CFD) and volume of fluid (VOF) method are adopted to investigate the flow behavior of melt and the cross-section of strand. The porosity characteristics, surface roughness and anisotropic mechanical properties of printed part are researched. The results show that, when the layer thickness is less than the diameter of nozzle, the squeezing of nozzle bottom face causes the melt climbing at the rear side of nozzle, which forms the bulges at the top corners of strand. Lower layer thickness helps to decrease the porosity and surface roughness. The proper increment of extrusion ratio can effectively reduce the porosity and improve the mechanical properties of the structures, yet large extrusion ratio may cause excessive squeezing of strand melt which will worsen the surface roughness and mechanical properties. This research contributes to a deeper understanding of FDM technology and provides guidance for the process optimization and structure design.

In this letter, a broadband low-profile dual circularly polarized reflectarray (dual-CP RA) for Ku-band satellite communications is proposed. A novel single-layer metasurface unit cell consisting of a functional layer, an air layer and a metal plate is investigated first. The functional layer is a metal structure printed on the F4B substrate. The air layer can effectively extend the bandwidth, and the overall profile is only 0.12 λ 0 , where λ 0 represents the wavelength at 11.725 GHz. To independently control the phase of left-handed circularly polarized (LHCP) and right-handed circularly polarized (RHCP) waves, Dynamic phase and Berry phase methods are employed by either changing the size of microstrip lines or the rotating of the cells. Finally, a dual-CP RA with 1600 cells is designed to realize two beams at 20° for LHCP wave and − 20° for RHCP wave at 11.725 GHz. The measured gain for LHCP wave is 29.1 dB with the aperture efficiency (AE) of 47% and 1-dB gain bandwidth of 37.4%, while the gain, AE, and bandwidth for RHCP wave are 29.22 dB, 48.3% and 37% respectively.

As one of the means of informal environmental regulation, environmental information disclosure has an essential impact on improving green energy efficiency. This paper selects the panel data of 113 environmental information disclosure cities in China from 2008 to 2018 and uses the Super-efficiency SBM model with undesirable outputs to measure green energy efficiency. Based on the measurement results, this paper empirically studies the impact of environmental information disclosure on green energy efficiency and its spatial spillover effect using the spatial Durbin model. The main conclusions are as follows: 1) From 2008 to 2018, the average green energy efficiency of 113 environmental information disclosure cities in China was 0.6676, and the regional distribution showed the characteristics of “high in the East and low in the west.” 2) Both environmental information disclosure and green energy efficiency have significant spatial correlation and show the characteristics of “high-high” and “low-low” agglomeration in spatial distribution. 3) Environmental information disclosure can significantly improve green energy efficiency in the region and surrounding areas. After the robustness test and endogenous test, the conclusion is still robust. 4) The impact of environmental information disclosure on green energy efficiency in the eastern region is significantly more significant than in the central and western regions. This paper provides a theoretical reference for the government to formulate corresponding environmental policies to promote green energy efficiency and promote green and sustainable economic development.

Significant wave height (SWH) is a key parameter for monitoring the state of waves. Accurate and long-term SWH forecasting is significant to maritime shipping and coastal engineering. This study proposes a transformer model based on an attention mechanism to achieve the forecasting of SWHs. The transformer model can capture the contextual information and dependencies between sequences and achieves continuous time series forecasting. Wave scale classification is carried out according to the forecasting results, and the results are compared with gated recurrent unit (GRU) and long short-term memory (LSTM) machine-learning models and the key laboratory of MArine Science and NUmerical Modeling (MASNUM) numerical wave model. The results show that the machine-learning models outperform the MASNUM within 72 h, with the transformer being the best model. For continuous 12 h, 24 h, 36 h, 48 h, 72 h, and 96 h forecasting, the average mean absolute errors (MAEs) of the test sets were, respectively, 0.139 m, 0.186 m, 0.223 m, 0.254 m, 0.302 m, and 0.329 m, and the wave scale classification accuracies were, respectively, 91.1%, 99.4%, 86%, 83.3%, 78.9%, and 77.5%. The experimental results validate that the transformer model can achieve continuous and accurate SWH forecasting, as well as accurate wave scale classification and early warning of waves, providing technical support for wave monitoring.