Corporate performance is the key indicator of availing the economic performances in all economies. Especially for the emerging economy, it is the oxygen for smooth economic operations. The study aims to investigate the influence of board characteristics on the corporate performance of the listed pharmaceuticals and chemicals sector from a developing country, namely Bangladesh. This empirical study examines eight attributes of the board and four financial performance indicators of the businesses. Here, the annual reports of the DSE-listed pharmaceutical and chemicals companies are considered to examine the impact of board attributes on corporate performance. Based on panel data analysis, this empirical study concludes that the fixed effect regression model is suitable for all four models. Except board size, the results demonstrate that all board attributes are generally statistically significant. Furthermore, it confirms that all the significant characteristics of the board are positively associated with corporate performance, except for board independence. The research offers valuable insights for policymakers, investors, organizations, and scholars, promoting optimal board structures, innovative solutions, and an enhanced understanding of corporate governance matters. This research explores the challenges in board attributes, which enhances our understanding of corporate governance matters and their impact over the last decade in the listed pharmaceutical and chemicals sectors in Bangladesh.

In this paper, we modeled and simulated two tandem solar cell structures (a) and (b), in a two-terminal configuration based on inorganic and lead-free absorber materials. The structures are composed of sub-cells already studied in our previous work, where we simulated the impact of defect density and recombination rate at the interfaces, as well as that of the thicknesses of the charge transport and absorber layers, on the photovoltaic performance. We also studied the performance resulting from the use of different materials for the electron and hole transport layers. The two structures studied include a bottom cell based on the perovskite material CsSnI₃ with a band gap energy of 1.3 eV and a thickness of 1.5 µm. The first structure has an upper sub-cell based on the CsSnGeI₃ material with an energy of 1.5 eV, while the second has an upper sub-cell made of Cs₂TiBr₆ with a band gap energy of 1.6 eV. The theoretical model used to evaluate the photocurrent density, current-voltage characteristic, and photovoltaic parameters of the constituent sub-cells and the tandem device was described. Current matching analysis was performed to find the ideal combination of absorber thicknesses that allows the same current density to be shared. An efficiency of 29.8% was obtained with a short circuit current density J_sc = 19.92 mA/cm², an open circuit potential V_oc = 1.46 V and a form factor FF = 91.5% with the first structure (a), for a top absorber thickness of CsSnGeI₃ of 190 nm, while an efficiency of 26.8% with J_sc = 16.74, V_oc = 1.50 V and FF = 91.4% was obtained with the second structure (b), for a top absorber thickness of Cs₂TiBr₆ of 300 nm. The objective of this study is to develop efficient, low-cost, stable and non-toxic tandem devices based on lead-free and inorganic perovskite.

The development of the maize agribusiness system is highly dependent on the role of social capital in facilitating interaction among actors in the chain of activities ranging from the provision of farm supplies to marketing. Therefore, this research aimed to characterize the key elements of social capital specifically bonding, bridging, and linking, as well as to demonstrate their respective roles. Data were collected from farmers and non-farmers actors engaged in various activities in the maize agribusiness system. The data obtained were processed using ATLAS Ti, applying open, axial, and selective coding techniques. The results showed the roles played by bonding, bridging, and linking social capital in the interaction between farmers and multiple actors in activities such as providing farm supplies, farming production, harvesting, post-harvest, and marketing. The combination of these social capital forms acted as the glue and wires that facilitated access to resources, collective decision-making, and reduced transaction costs. These results have theoretical implications, suggesting that bonding, bridging, and linking should be combined with the appropriate role composition for each activity in the agribusiness system.

Farm households in developing countries are often involved in a variety of livelihood income-generating activities to achieve basic needs and enhance food security. However, little attention has been given to investigating the effect of livelihood diversification strategies on the adoption of agricultural land management practices. This study explored the nexus between livelihood diversification and Agricultural Land Management (ALM) practices in the Southern Ethiopian Highlands. Data for this study were gathered through a structured questionnaire, interviews, and focus group discussions. A total of 423 sample respondents were selected by using multistage random sampling techniques. The data were analyzed using the Inverse Herfindahl Hirschman Diversity Index (IHHDI), the multinomial logit model (MNL), and the probit regression model. The findings of the study revealed that on-farm income activities are the most dominant livelihood income strategies (69.1%), followed by non-farm (21%) and off-farm (9.64%). The multinomial logit model analysis demonstrated that variables such as sex, education, family size, distance to market, land size, extension contact, membership in cooperatives, and household income were the major drivers of farmers income diversification activities (p<0.05). The results of the probit analysis indicated that income from crop production, daily labor work, rents from farmland, and farm assets have a positive and significant effect on households' decisions to implement ALM practices. In contrast, incomes from remittance and migrant sources have a negative but statistically significant impact on the adoption of ALM measures. The farm household sources of income-generating strategies substantially affected the adoption intensity of ALM measures. Income generated from the on-farm sector alone cannot be considered a core income-generating activity for households or a means of achieving food security. Therefore, land management policies and program implementations should consider farmers’ livelihood diversification and income-generating strategies. In addition, such interventions need to promote sustainable farming practices, enhance innovation, and related measures for the adoption of ALM measures to ensure land sustainability.

Brain tumors are a primary factor causing cancer-related deaths globally, and their classification remains a significant research challenge due to the variability in tumor intensity, size, and shape, as well as the similar appearances of different tumor types. Accurate differentiation is further complicated by these factors, making diagnosis difficult even with advanced imaging techniques such as magnetic resonance imaging (MRI). Recent techniques in artificial intelligence (AI), in particular deep learning (DL), have improved the speed and accuracy of medical image analysis, but they still face challenges like overfitting and the need for large annotated datasets. This study addresses these challenges by presenting two approaches for brain tumor classification using MRI images. The first approach involves fine-tuning transfer learning cutting-edge models, including SEResNet, ConvNeXtBase, and ResNet101V2, with global average pooling 2D and dropout layers to minimize overfitting and reduce the need for extensive preprocessing. The second approach leverages the Vision Transformer (ViT), optimized with the AdamW optimizer and extensive data augmentation. Experiments on the BT-Large-4C dataset demonstrate that SEResNet achieves the highest accuracy of 97.96%, surpassing ViT’s 95.4%. These results suggest that fine-tuning and transfer learning models are more effective at addressing the challenges of overfitting and dataset limitations, ultimately outperforming the Vision Transformer and existing state-of-the-art techniques in brain tumor classification.