A machine learning ensemble approach for predicting solar-sensitive hybrid photocatalysts on hydrogen evolution

Physica Scripta, Journal Year: 2024, Volume and Issue: 99(7), P. 076015 - 076015

Published: June 10, 2024

Abstract Hydrogen, as the lightest and most abundant element in universe, has emerged a pivotal player quest for sustainable energy solutions. Its remarkable properties, such high density zero emissions upon combustion, make it promising candidate addressing pressing challenges of climate change transitioning towards clean renewable future. In an effort to improve efficiency reduce experimental costs, we adopted machine learning techniques this study. Our focus turned predictive analyses hydrogen evolution values using three photocatalysts, namely, graphene-supported LaFeO 3 (GLFO), LaRuO (GLRO), BiFeO (GBFO), examining their correlation with varying levels pH, catalyst amount, H 2 O concentration. To achieve this, diverse range models are used, including Random Forest (RF), Decision Tree (DT), Support Vector Machine (SVM), XGBoost, Gradient Boosting, AdaBoost—each bringing its strengths modeling arena. An important step involved combining effective models—Random Forests, XGBoost—into ensemble model. This collaborative approach aimed leverage collective overall predictability. The model powerful tool understanding photocatalytic evolution. Standard metrics were employed assess performance our prediction model, encompassing R squared, Root Mean Squared Error (RMSE), (MSE), Absolute (MAE). yielded results showcase exceptional accuracy, squared 96.9%, 99.3%, 98% GLFO, GBFO, GLRO, respectively. Moreover, demonstrates minimal error rates across all metrics, underscoring robust capabilities highlighting efficacy accurately forecasting intricate relationships between GLRO influencing factors.

Language: Английский

---

VoteDroid: a new ensemble voting classifier for malware detection based on fine-tuned deep learning models

Multimedia Tools and Applications, Journal Year: 2024, Volume and Issue: unknown

Published: May 18, 2024

Abstract In this work, VoteDroid a novel fine-tuned deep learning models-based ensemble voting classifier has been proposed for detecting malicious behavior in Android applications. To end, we adopting the random search optimization algorithm deciding structure of models used as voter classifiers classifier. We specified potential components that can be each model and left taking decision about including number component should its location structure. This method to build three different namely CNN-ANN, pure CNN, ANN. After selecting best DL model, selected have trained tested using constructed image dataset. Afterward, suggested hybridizing deep-learning form one with two working modes MMR (Malware Minority Rule) LMR (Label Majority Rule). our knowledge, is first time an hybridized way malware detection. The results showed were promising, where classification accuracy exceeded 97% all experiments.

Language: Английский

---

Stacked machine learning approach for predicting evolved hydrogen from sugar industry wastewater

International Journal of Hydrogen Energy, Journal Year: 2024, Volume and Issue: 85, P. 75 - 87

Published: Aug. 24, 2024

Language: Английский

---

Integrating Experimental and Machine Learning Approaches for Predictive Analysis of Photocatalytic Hydrogen Evolution Using Cu/g-C3N4

Renewable Energy, Journal Year: 2024, Volume and Issue: unknown, P. 121737 - 121737

Published: Oct. 1, 2024

Language: Английский

---

Django-based framework database for leakage detection using machine learning for water distribution networks

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 149, P. 110525 - 110525

Published: March 15, 2025

Language: Английский

---

Fractional data driven controller based on adaptive neural network optimizer

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 257, P. 125077 - 125077

Published: Aug. 10, 2024

Language: Английский

---

Research on hot deformation behavior of GH98 superalloy under various stress conditions based on a deep learning approach

Journal of materials research/Pratt's guide to venture capital sources, Journal Year: 2024, Volume and Issue: 39(21), P. 3007 - 3017

Published: Sept. 23, 2024

Language: Английский

---