The issue of employability has gained significant importance, not only for graduate students but also for higher educational institutions. In this regard, employability prediction models using machine learning have emerged as crucial techniques for assessing students' potential to secure employment after graduation. Enhancing university graduate employability is critical because student unemployment is a global concern that has widespread negative effects on both individuals and institutions. Therefore, focusing on graduate employability predictions using machine learning techniques is considered essential in addressing this issue. Traditionally, demographic and academic attributes, such as CGPA, have been considered key factors in determining student employment status. However, research suggests that various other factors, such as student satisfaction, might influence employability. This study employs machine learning techniques to identify the factors that affect graduate student employability. The objective is to investigate the features significantly influencing students' ability to secure employment. Data was collected from Malaysia's Ministry of Education's graduate tracer study (SKPG). Several classification algorithms were applied, including Logistic Regression, Random Forest, Naïve Bayes, Support Vector Machine, Extreme Gradient Boosting, and Artificial Neural Networks (ANN). The results show that ANN achieved the highest accuracy, with around 80%. The findings also revealed that student demographic and academic features and student satisfaction level with the university facilities (e.g., library and counseling service) are considered significant for graduate student employability predictions. Consequently, the empirical results can help higher educational institutions enhance facilities and prepare students with the necessary skills for future employability.

Rapidly growing medical imaging volumes have led to an increasing workload for radiologists, creating the need for automated tools that can support interpretation and reduce reporting delays. Vision-language models (VLMs) can generate clinically relevant captions to accelerate report drafting, but their varying parameter scales require evaluation for clinical utility. This study evaluated fine-tuned VLMs on the Radiology Objects in Context version 2 (ROCOv2) dataset containing 116,635 images across multiple modalities. We compared four Large VLMs (LVLMs) (LLaVA-Mistral-7B, LLaVA-Vicuna-7B, LLaVA-1.5-LLaMA-7B, IDEFICS-9B) against four Smaller VLMs (SVLMs) (MoonDream2, Qwen 2-VL, Qwen-2.5, SmolVLM) alongside two fully fine-tuned baseline architectures (VisionGPT2 and CNN-Transformer). Low-Rank Adaptation (LoRA), applied to fewer than 1% of selected model parameters, proved the optimal performance among adaptation strategies, outperforming broader LoRA configurations that we evaluated. LLaVA-Mistral-7B achieved the highest performance (Relevance: 0.516, Factuality: 0.118), substantially exceeding VisionGPT2 baseline (0.325, 0.028). Among SVLMs, MoonDream2 reached a relevance score of 0.466, surpassing LLaVA-1.5 (0.462) despite using approximately 74% fewer parameters. Models showed distinct performance hierarchy with LVLMs (0.273–0.317 overall), SVLMs (0.188–0.279), and baselines (0.154–0.177). To investigate performance enhancement strategies, we prepended ResNet-50-predicted image modality labels at inference time for underperforming SVLMs. This intervention produced variable results, with SmolVLM showing marginal improvement, Qwen-2.5 gaining 6.4%, but Qwen 2-VL experiencing a 21.6% performance reduction. Our results provide quantitative guidance for VLM selection in medical imaging. Although model size strongly influences performance, the findings indicate that architectural design and lightweight adaptation can enable select small models to achieve viable performance for resource-constrained screening scenarios.

Rapidly growing medical imaging volumes have increased radiologist workloads, creating demand for automated tools that support interpretation and reduce reporting delays. Vision-language models (VLMs) can generate clinically relevant captions to accelerate report drafting, yet their varying parameter scales require systematic evaluation for clinical utility. This study evaluated ten multimodal models fine-tuned on the Radiology Objects in Context version 2 (ROCOv2) dataset containing 116,635 images across eight modalities. We compared four Large VLMs (LVLMs) including LLaVA variants and IDEFICS-9B against four Small VLMs (SVLMs) including MoonDream2, Qwen variants, and SmolVLM, alongside two fully fine-tuned baseline architectures (VisionGPT2 and CNN-Transformer). Low-Rank Adaptation (LoRA), applied to fewer than 1% of selected model parameters, proved optimal among adaptation strategies, outperforming broader LoRA configurations. Models were assessed on relevance (semantic similarity) and factuality (concept-level correctness) metrics. Performance showed clear stratification: LVLMs (0.273 to 0.317 overall), SVLMs (0.188 to 0.279), and baselines (0.154 to 0.177). LLaVA-Mistral-7B achieved the highest performance with relevance and factuality scores of 0.516 and 0.118, respectively, substantially exceeding the VisionGPT2 baseline (0.325, 0.028). Among the SVLMs, MoonDream2 demonstrated competitive relevance (0.466), approaching the performance of some LVLMs despite its smaller size. To investigate performance enhancement strategies for underperforming SVLMs, we prepended predicted imaging modality labels at inference time, which yielded variable results. These findings provide quantitative benchmarks for VLM selection in medical imaging, demonstrating that while model scale influences performance, architectural design and targeted adaptation enable select compact models to achieve competitive results.

More than 48 million people have been affected by diarrhea and nearly 1 million people have been affected by malaria due to climatic variability in the Sylhet region from 1988-2005. Therefore, the aim of this study is to seek the impact of climate variability on human health. Two types of data including monthly climatic data for six years collected from Bangladesh Meteorological Department, Sylhet and outpatient visit data for six years from two hospitals have collected for this study. Inferential statistics i.e. Poisson Regression is used to analyze data. The analysis has been done using STATA 12 software; temperature, rainfall, humidity are the independent variable and number of patient is dependent variable. It is found that temperature, rainfall and humidity has significance influence on diarrhea, skin disease and mental disorder. Temperature changes of one degree and rainfall changes of 1 mm may increase 2.5%, 0.19% diarrhea patient respectively. Similarly, temperature changes of one degree and humidity changes of 1% change may increase 6.2% and 2% mental disorder patient respectively. In the same way temperature changes of one degree and humidity changes of 1% change may increase 2.5% and 2% skin diseases patient respectively. The significance level of temperature, rainfall and humidity for diarrhea, skin diseases and mental disorder is 0.00. The present study finds that diarrhea, mental disorder and skin disease are influenced by climate variability and seasonal change. The impact will likely vary by region, the sensitivity of populations. Poor people are more vulnerable than rich people in case of climate change. This study also finds that the health sector has not been developed yet, and the related data is unavailable for research work. Therefore, the present study suggests that it is important to develop a database of diseases for further research.