High-performance glass classification using advanced machine learning and deep learning algorithms with a comprehensive feature analysis DOI Creative Commons

Mohammed Bouziane,

Abdelghani Bouziane, Samia Larguech

et al.

AIP Advances, Journal Year: 2025, Volume and Issue: 15(5)

Published: May 1, 2025

Glass classification with accuracy is highly required in construction, automotive, and electronics industries, where material properties like transparency strength are vital. Traditional practices, though effective, time-consuming non-scalable. This paper proposes a solution based on Machine Learning Deep to automate scale up the of glass classification. The work uses dataset 214 samples nine chemical physical properties. Exploratory Data Analysis provides significant patterns verifies pre-determined classes through clustering techniques Gaussian Mixture Models. Advanced learning algorithms Random Forest (RF), XGBoost, Support Vector Machines, Bidirectional Long Short-Term Memory (BiLSTM) networks applied for Findings prove RF XGBoost provide highest accuracy, BiLSTM be best recognizing complex data patterns. Feature importance analysis pinpoints features identifies magnesium barium among those used distinguish between types. detailed evaluation highlights potential AI-based methods revolutionize classifying increased efficacy, details.

Language: Английский

Regulating emission in Er doped silicate glass and fiber via coordination engineering DOI Creative Commons
Yuqing Li, Yan Sun, Xin Wang

et al.

Journal of Materiomics, Journal Year: 2025, Volume and Issue: unknown, P. 101057 - 101057

Published: April 1, 2025

Language: Английский

Citations

0
High-performance glass classification using advanced machine learning and deep learning algorithms with a comprehensive feature analysis DOI Creative Commons

Mohammed Bouziane,

Abdelghani Bouziane, Samia Larguech

et al.

AIP Advances, Journal Year: 2025, Volume and Issue: 15(5)

Published: May 1, 2025

Glass classification with accuracy is highly required in construction, automotive, and electronics industries, where material properties like transparency strength are vital. Traditional practices, though effective, time-consuming non-scalable. This paper proposes a solution based on Machine Learning Deep to automate scale up the of glass classification. The work uses dataset 214 samples nine chemical physical properties. Exploratory Data Analysis provides significant patterns verifies pre-determined classes through clustering techniques Gaussian Mixture Models. Advanced learning algorithms Random Forest (RF), XGBoost, Support Vector Machines, Bidirectional Long Short-Term Memory (BiLSTM) networks applied for Findings prove RF XGBoost provide highest accuracy, BiLSTM be best recognizing complex data patterns. Feature importance analysis pinpoints features identifies magnesium barium among those used distinguish between types. detailed evaluation highlights potential AI-based methods revolutionize classifying increased efficacy, details.

Language: Английский

Citations

0