6277 research outputs found
Sort by
Exploratory data analysis and data visualization on accidental drug related deaths
Get PDFThe drug overdose epidemic in the United States is rapidly getting worse with substantial associated public health effects. Covering 10,654 cases over a decade (2012–2022), this study analyses an extensive dataset of accidental drug-related deaths in Connecticut. We process and analyze this data using Python and Tableau, we then use LSTM to predict how many people will show up in the designated intervals. The ages of individuals were stated as a mean value 43.52 (SD =12.60) years with a range between 13 and 87 years, bimodally distributed around the mid-30s to mid-50's. Overall, 74.14% were male and 85.48 % white in race/ethnicity. Significant increases were seen in accidental drug-related deaths. The most implicated substances were any opioids, fentanyl (alone or in combination), cocaine alone, heroin and ethanol. Crucially, some 89.05% of the cases had co-abuse with multiple drugs by one person who showed evidence that poly-substance use is commonplace in this community. Most deaths involved fentanyl (309, with a mode at 36 years (229 out of 309 deaths) and r (0.51) associated with ‘any opioid’, the primary cause of death). New Haven, Hartford and Fairfield counties stood out as hotspots for overdoses in geographic analysis. The LSTM model achieved a Root Mean Square Error (RMSE) of 7.25 and a Mean Absolute Error (MAE) of 5.62, predicting a sustained annual increase in deaths over the next three years. This federal and state partnership provides a model for using existing surveillance resources to inform targeted overdose intervention strategies, with an emphasis on the rise of fentanyl positivity among decedents
Diabetic patient readmission predictive analysis: a comparative study of machine learning models of hospital readmissions
Get PDFHospital readmissions are a major concern in healthcare due to their impact on patient morbidity, hospital workload, and increased healthcare costs. This study aims to predict the readmission of diabetic patients within 30 days postdischarge, addressing the critical need for efficient resource allocation and improved patient care. The motivation stems from the high prevalence of diabetes and the associated risk of complications leading to frequent readmissions. Using a comprehensive dataset from 1999 to 2008, encompassing 130 US healthcare facilities, the research employs advanced machine learning techniques to develop a predictive model. Data preprocessing and feature engineering are meticulously applied to enhance model accuracy. Various classifiers, including Decision Tree, K-Nearest Neighbors, AdaBoost, Naive Bayes, Random Forest, and Logistic Regression, are evaluated against standard performance metrics. Results indicate that the Naive Bayes model achieves the highest F1 score of 0.81, outperforming other models. The study concludes that predictive modeling can significantly enhance clinical decision making and optimize healthcare resources, underlining its importance in managing diabetic patient care
"Where do we belong?" Collaborative insights from RAISE Special Issue Groups' (Early Careers and Research Evaluation) Writing Project
Get PDF"Who am I? Navigating professional identity as an ethnic minority early career academic
Get PDFFinding the answer to my academic identity and a sense of belonging is something I have questioned since I just started my doctorate journey. It has been a question lingering in my mind with no clear answer where one will go after a few years of completing a PhD. Yet, without a doubt, the challenges I experienced demonstrate that constructing my identity in my discipline and the wider sector will be much more challenging than one would have hoped. This reflection article aims to share a personal account of a doctoral researcher transitioning to an independent professional. Key aspects, such as their experience constructing their identity in their field, the professional developmental support as a doctoral researcher in preparation and during that transition, and the challenges with employment opportunities for the wider doctoral graduate community will be discussed. In addition to drawing on previous literature, provocative questions will be shared with the higher education sector to consider what more we can do to support the lives of postgraduate researchers after their doctorate to support their sense of belonging. Championing that early career professionals need to have the graduates' attributes to make it in the "real world" is one thing. But the support to meet individual needs is another. Having the right skills and experiences is difficult and valuable. Yet, do we know if our graduates are fully ready to become independent professionals in the real world
Machine learning approaches for accurate energy content prediction in foods using nutritional data
Get PDFThis study marks a step toward more effectively translating nutritional information to inform public health policy as well as individual dietary choices. Motivated by the increase in diet-related health issues, this research aims to analyze a comprehensive nutritional dataset to uncover valuable insights. Using the USDA National Nutrient Database, the study employs data preprocessing to clean the data, exploratory data analysis to identify hidden patterns, and various machine learning models to predict nutritional values. The results demonstrate the usefulness of these models in explaining the composition data and highlight a range of trends and relationships within the observed amounts. The discussion emphasizes that these findings could be instrumental in guiding health professionals and policymakers toward healthier dietary guidelines. The significance of this research lies in its potential to advance public health through more sophisticated nutritional recommendations
metainc: Assessment of inconsistency in meta-analysis using decision thresholds
No full textAssessment of inconsistency in meta-analysis by calculating the Decision Inconsistency index (DI) and the Across-Studies Inconsistency (ASI) index. These indices quantify inconsistency taking into account outcome-level decision thresholds
Data sources:
Access Repository Dashboard
Do you manage Solent University Research Portal? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!