Prioritizing Patient Selection in Clinical Trials: A Machine Learning Algorithm for Dynamic Prediction of In-Hospital Mortality for ICU Admitted Patients Using Repeated Measurement Data

Journal of Clinical Medicine, Journal Year: 2025, Volume and Issue: 14(2), P. 612 - 612

Published: Jan. 18, 2025

Background: A machine learning prognostic mortality scoring system was developed to address challenges in patient selection for clinical trials within the Intensive Care Unit (ICU) environment. The algorithm incorporates Red blood cell Distribution Width (RDW) data and other demographic characteristics predict ICU alongside existing systems like Simplified Acute Physiology Score (SAPS). Methods: algorithm, defined as a Mixed-effects logistic Random Forest binary (MixRFb), integrates (RF) classification with mixed-effects model outcomes, accounting repeated measurement data. Performance comparisons were conducted RF proposed MixRFb algorithms based solely on SAPS scoring, additional evaluation using descriptive receiver operating characteristic curve incorporating RDW’s predictive ability. Results: MixRFb, RDW covariates, outperforms SAPS-based variant, achieving an area under of 0.882 compared 0.814. Age identified most significant predictors mortality, reported by variable importance plot analysis. Conclusions: demonstrates superior efficacy predicting in-hospital identifies age primary predictors. Implementation this could facilitate trials, thereby improving trial outcomes strengthening ethical standards. Future research should focus enriching robustness, expanding its applicability across diverse settings demographics, integrating markers improve capabilities.

Language: Английский

---

A real-time prediction method for rate of penetration sequence in offshore deep wells drilling based on attention mechanism-enhanced BiLSTM model

Ocean Engineering, Journal Year: 2025, Volume and Issue: 325, P. 120820 - 120820

Published: March 2, 2025

Language: Английский

---

ISI Net: A novel paradigm integrating interpretability and intelligent selection in ensemble learning for accurate wind power forecasting

Energy Conversion and Management, Journal Year: 2025, Volume and Issue: 332, P. 119752 - 119752

Published: April 2, 2025

Language: Английский

---

Improving the Accuracy of Groundwater Level Forecasting by Coupling Ensemble Machine Learning Model and Coronavirus Herd Immunity Optimizer

Water Resources Management, Journal Year: 2025, Volume and Issue: unknown

Published: May 3, 2025

Language: Английский

---

Higher order Weighted Random <i>k</i> Satisfiability ($k = 1, 3$) in Discrete Hopfield Neural Network

AIMS Mathematics, Journal Year: 2025, Volume and Issue: 10(1), P. 159 - 194

Published: Jan. 1, 2025

<p>Researchers have explored various non-systematic satisfiability approaches to enhance the interpretability of Discrete Hopfield Neural Networks. A flexible framework for has been developed investigate diverse logical structures across dimensions and improved lack neuron variation. However, logic phase this approach tends overlook distribution characteristics literal states, ratio negative literals not mentioned with higher-order clauses. In paper, we propose a new named Weighted Random $k$ Satisfiability ($k = 1, 3$), which implements in The proposed logic, integrated into Network, established structure by incorporating during phase. This enhancement increased network's storage capacity, improving its ability handle complex, high-dimensional problems. advanced was evaluated learning metrics. When values were $r 0.2$, 0.4, 0.6, 0.8, demonstrated potential better performances smaller errors. Furthermore, performance positive impact on management synaptic weights. results indicated that optimal global minimum solutions are achieved when set 0.8$. Compared state-of-the-art structures, novel more significant achieving solutions, particularly terms literals.</p>

Language: Английский

---

Enhancing UAV Security Against GPS Spoofing Attacks Through a Genetic Algorithm-Driven Deep Learning Framework

Information, Journal Year: 2025, Volume and Issue: 16(2), P. 115 - 115

Published: Feb. 7, 2025

Unmanned Aerial Vehicles (UAVs) are increasingly employed across various domains, including communication, military, and delivery operations. Their reliance on the Global Positioning System (GPS) renders them vulnerable to GPS spoofing attacks, in which adversaries transmit false signals manipulate UAVs’ navigation, potentially leading severe security risks. This paper presents an enhanced integration of Long Short-Term Memory (LSTM) networks with a Genetic Algorithm (GA) for detection. Although GA–neural network combinations have existed decades, our method expands GA’s search space optimize wider range hyperparameters, thereby improving adaptability dynamic operational scenarios. The framework is evaluated using real-world dataset that includes authentic malicious under multiple attack conditions. While we discuss strategies mitigating CPU resource demands computational overhead, acknowledge direct measurements energy consumption or inference latency not included present work. Experimental results show proposed LSTM–GA approach achieved notable increase classification accuracy (from 88.42% 93.12%) F1 score 87.63% 93.39%). These findings highlight system’s potential strengthen UAV against provided hardware constraints other limitations carefully managed real deployments.

Language: Английский

---

An Integrated Approach Using Lars-Wg and Deep Learning for River Flow Prediction in Diverse Regions

Published: Jan. 1, 2025

Language: Английский

---

Prediction of Long-Period Ground Motion Responses for High-Rise Buildings Using Physics-Assisted Fully Convolutional Neural Network

Journal of Building Engineering, Journal Year: 2025, Volume and Issue: unknown, P. 112264 - 112264

Published: March 1, 2025

Language: Английский

---

Stock Market Index Prediction Using CEEMDAN‐LSTM‐BPNN‐Decomposition Ensemble Model

Journal of Applied Mathematics, Journal Year: 2025, Volume and Issue: 2025(1)

Published: Jan. 1, 2025

This study investigates the forecasting of Deutscher Aktienindex (DAX) market index by addressing nonlinear and nonstationary nature financial time series data using CEEMDAN decomposition method. The technique is used to decompose into intrinsic mode functions (IMFs) residuals, which are classified low‐frequency (LF), medium‐frequency (MF), high‐frequency (HF) components. Long short‐term memory (LSTM) networks applied MF HF components, while backpropagation neural network (BPNN) utilized for LF resulting in a robust hybrid model termed CEEMDAN‐LSTM‐BPNN. To evaluate performance proposed model, we compare it against several benchmark models, including ARIMA, RNN, LSTM, GRU, BIGRU, BILSTM, BPNN, CEEMDAN‐LSTM, CEEMDAN‐GRU, CEEMDAN‐BPNN, CEEMDAN‐GRU‐BPNN, across different training–testing splits (70% training/30% testing, 80% training/20% 90% training/10% testing). model’s predictive accuracy measured six metrics: root mean squared error (RMSE), absolute (MAE), percentage (MAPE), symmetric (SMAPE), logarithmic (RMSLE), R ‐squared. further assess performance, conduct Diebold–Mariano (DM) test forecast between models confidence set (MCS) statistical significance improvement. results demonstrate that CEEMDAN‐LSTM‐BPNN significantly outperforms other methods terms accuracy, with DM MCS tests confirming superiority multiple evaluation metrics. findings highlight importance combining advanced deep learning forecasting. research contributes development more accurate techniques, offering valuable implications decision‐making risk management.

Language: Английский

---

Depth determination of simulated biological tissue using X-ray radiography and feature extraction techniques: Evaluation with Bi-LSTM neural network

Journal of Radiation Research and Applied Sciences, Journal Year: 2025, Volume and Issue: 18(2), P. 101406 - 101406

Published: March 8, 2025

Language: Английский

---