Cancer Detection Using a New Hybrid Method Based on Pattern Recognition in MicroRNAs Combining Particle Swarm Optimization Algorithm and Artificial Neural Network

Big Data and Cognitive Computing, Journal Year: 2024, Volume and Issue: 8(3), P. 33 - 33

Published: March 19, 2024

MicroRNAs (miRNAs) play a crucial role in cancer development, but not all miRNAs are equally significant detection. Traditional methods face challenges effectively identifying cancer-associated due to data complexity and volume. This study introduces novel, feature-based technique for detecting attributes related cancer-affecting microRNAs. It aims enhance diagnosis accuracy by the most relevant various types using hybrid approach. In particular, we used combination of particle swarm optimization (PSO) artificial neural networks (ANNs) this purpose. PSO was employed feature selection, focusing on informative miRNAs, while ANNs were recognizing patterns within miRNA data. method overcome limitations traditional analysis reducing redundancy key genetic markers. The application showed improvement detection cancers, including breast lung melanoma. Our approach demonstrated higher precision compared existing methods, as evidenced different datasets. concludes that integration provides more efficient, cost-effective, accurate via analysis. can serve supplementary tool potentially aid developing personalized treatments.

Language: Английский

---

Ensemble Technique-Based Short-Term Supply and Demand Forecasting with Features Selection Approach in Decentralized Energy Systems

Published: Dec. 31, 2024

Article Ensemble Technique-Based Short-Term Supply and Demand Forecasting with Features Selection Approach in Decentralized Energy Systems Adugna Gebrie Jember 1, Ruiyu Bao Zijia Yao Zhao Wang 1,*, Zhenyu Zhou 1 Xiaoyan 2 State Key Laboratory of Alternate Electrical Power System Renewable Sources, North China Electric University, Beijing 102206, Graduate School Science Engineering, Ibaraki Mito 310-8512, Japan * Correspondence: [email protected] Received: 21 August 2024; Revised: December Accepted: Published: 31 2024 Abstract: energy systems (DESs) present a paradigm shift toward more sustainable resilient electricity networks the increasing integration renewable sources. Accurately forecasting supply demand is crucial for efficient operation DES. This paper focuses on accurate short-term using machine learning (ML) algorithms across different seasons within small-scale DES, including photovoltaic (PV) generation, wind load demand. Initially, we develop multiple base model ML long memory (LSTM), convolutional neural (CNN), eXtreme gradient boosting (XGBoost), recurrent (RNN) feature selection approaches to improve accuracy reduce complexity. These models leverage key temporal spatial features seasonal variations process overfitting seasons. To further forecast errors, propose robust ensemble techniques simple averaging (SA), weighted (WA), Stacking. In algorithms, technique combines models’ forecasts produce final by leveraging strengths compensating weaknesses individual models. Finally, numerical simulations are conducted Python, Keras, TensorFlow libraries develop, train, evaluate, validate effectiveness developed proposed techniques. The results demonstrate that approach offers solution problems work both novel effective from perspectives application, combination, performance improvement.

Language: Английский

---

SHAP-Guided GWO and ABC Feature Selection for Early Diabetes Prediction with XGBoost

Published: May 2, 2024

The study examines the relationship of autonomous motivation with physical activity in type 2 diabetes mellitus (T2DM) patients after participation a fitness intervention trial. also longitudinally explores various factors regarding newly diagnosed population, which gives useful insight into dietary self-care and treatment diabetes. This pointed towards research gaps that are existing hence further need conducting for gender-specific inequities, management cardio metabolic risk, provision autonomy support, psychological issues, foot health among diabetic patients. work hereby brings feature selection interpretable modelling techniques prediction using optimization algorithms such as Grey Wolf Optimization (GWO) Artificial Bee Colony (ABC), together SHAP-XGBoost. Concerning comparison between GWO SHAP-XGBoost ABC SHAP-XGBoost, it can be said latter is best performer. to exhibit 82.26% accuracy. outcomes this very improvement therapies, particular, tailored options overall

Language: Английский

---

An Optimization Method for Short-Term Power Load Forecasting Based on IPSO-SVM

Published: June 21, 2024

Language: Английский

---