Siamese Graph Convolutional Split-Attention Network with NLP based Social Sentimental Data for enhanced stock price predictions

Journal Of Big Data, Год журнала: 2024, Номер 11(1)

Опубликована: Окт. 29, 2024

Abstract Predicting stock market behavior using sentiment analysis has become increasingly popular, as customer responses on platforms like Twitter can influence trends. However, most existing sentiment-based models struggle with two major issues: inaccuracy and high complexity. These problems lead to frequent prediction errors make the difficult implement in real-time trading systems. To address these challenges, this paper proposes a new method called Siagra-ConSA-HSOA (Siamese Graph Convolutional Split-Attention Network NLP-based Social Sentiment Data). Two data sources feed model: specifically, NIFTY-50 Stock Market sentiment. Through Natural Language Processing (NLP), raw is pre-processed key features are extracted before they fused into unified dataset cross-domain transformer, namely CDSFT, then Circle-Inspired Optimization Algorithm (CIOA) selects important from dataset. This decreases complexity of model without losing essential information. Finally, (SGCSAN) for promisingly predicting whether prices going hit ground fly again or nosedive Humboldt Squid (HSOA) introduced further improve accuracy lesser error generation. The proposed achieved 99.9% 99.8% recall testing stage, meaning that such performs better than current approaches both efficiency. Thus, glimmer shall be able overcome some main techniques used market. GitHub Repository: https://github.com/jramans2/Siamese-GCN-SplitAttention-Stock-Prediction.git

Язык: Английский

Evaluation of serum transferrin microheterogeneity for the diagnosis of congenital N-glycosylation defects

SCT Proceedings in Interdisciplinary Insights and Innovations., Год журнала: 2024, Номер 3, С. 374 - 374

Опубликована: Дек. 31, 2024

Introduction: transferrin is a glycoprotein produced in the liver, whose function to transport iron tissues. It has been used mainly for differential diagnosis of anemias as biomarker. There are different isoforms due difference their glycosylation patterns. This microheterogeneity allowed its use biomarker Congenital Disorders Glycosylation; genetic diseases result mutations genes that encode enzymes post-translational mechanism protein glycosylation.Objective: evaluate serum Glycosylation CubaMethods: descriptive and cross-sectional study was developed at National Center Medical Genetics period from 2016 2022. The analytical method isoelectric focusing with immunofixation described by Van Eijik et al 1983. Serum samples 26 patients multisystem clinical symptoms suspicion having disease without definitive di-agnosis were processedResults: IEF us determine pattern Tf. An altered Tf found four samples, two type I II.Conclusions: glycoforms positive patients, thus demonstrating presence Protein N-glycosylation Cuba

Язык: Английский