Yu Song
Research Subjects
Yu Song has a total of 46 co-authors. Here are the top co-authors by number of shared publications:
- David A. Lange
University of Illinois Urbana-Champaign
5 shared publications - Mathieu Bauchy
Department of Civil and Environmental Engineering, University of California, Los Angeles, USA
5 shared publications Gaurav Sant
Laboratory for the Chemistry of Construction Materials (LC2) 5731J Boelter Hall, Department of Civil and Environmental Engineering, University of California, Los Angeles, CA 90095, USA
3 shared publications- Gaurav Sant
University of California, Los Angeles
3 shared publications Alex Ginella
Physics of AmoRphous and Inorganic Solids Laboratory (PARISlab) 5731B Boelter Hall, Department of Civil and Environmental Engineering, University of California, Los Angeles, CA 90095, USA
2 shared publicationsBenjamin Gallagher
Electric Power Research Institute (EPRI) 3420 Hillview Avenue, Palo Alto, CA 94304, USA
2 shared publications
Latest Publications
- Open Access
- 19 February 2025
- Published by Elsevier in Waste Management
- Vol. 196, 60-70
Classification and predictive leaching risk assessment of construction and demolition waste using multivariate statistical and machine learning analyses
- 19 February 2025
- Published by Elsevier in Waste Management
- Vol. 196, 60-70
Open AccessClassification and predictive leaching risk assessment of construction and demolition waste using multivariate statistical and machine learning analyses
- 2 January 2024
- Published by Royal Society of Chemistry (RSC) in Materials Horizons
- Vol. 11 (6) , 1448-1464
Predicting rare earth elements concentration in coal ashes with multi-task neural networks
Our multi-task neural network approach simultaneously predicts the concentration of all types of rare earth elements (REEs) in coal ashes, with an improved accuracy and robustness as compared to conventional single-task neural networks.
- 2 January 2024
- Published by Royal Society of Chemistry (RSC) in Materials Horizons
- Vol. 11 (6) , 1448-1464
Predicting rare earth elements concentration in coal ashes with multi-task neural networks
Our multi-task neural network approach simultaneously predicts the concentration of all types of rare earth elements (REEs) in coal ashes, with an improved accuracy and robustness as compared to conventional single-task neural networks.
- Open Access
- 22 December 2023
- Published by Elsevier in Cleaner Materials
- Vol. 11
Estimating attached mortar paste on the surface of recycled aggregates based on deep learning and mineralogical models
- 22 December 2023
- Published by Elsevier in Cleaner Materials
- Vol. 11
Open AccessEstimating attached mortar paste on the surface of recycled aggregates based on deep learning and mineralogical models
- 25 November 2023
- Published by Elsevier in Construction and Building Materials
- Vol. 411
Continuous monitoring of the moisture, shrinkage, and carbonation effects on foam concrete performance
- 25 November 2023
- Published by Elsevier in Construction and Building Materials
- Vol. 411
Continuous monitoring of the moisture, shrinkage, and carbonation effects on foam concrete performance
- Open Access
- 24 October 2022
- Published by Springer Nature in Research Square
Predicting Rare Earth Elements Concentration in Coal Ashes with Multi-Task Neural Networks
The increasing demand for rare earth elements (REEs) makes them a scarce strategic resource for technical developments. In that regard, harvesting REEs from coal ashes—a waste byproduct from coal power plants—offers an alternative solution to conventional ore-based extraction. However, this approach is bottlenecked by our ability to screen coal ashes bearing large concentrations of REEs from feedstocks—since measuring the REE content in ashes is a time-consuming and costly task requiring advanced analytical tools. Here, we propose a machine learning approach to predict the REE contents based on the bulk composition of coal ashes (which is easily measurable under the current testing protocol). We introduce a multi-task neural network that simultaneously predicts the contents of different REEs and, importantly, exhibits notably improved accuracy than the single-task models. Further model analyses reveal key data patterns for screening coal ashes with high REE concentrations.Teaser: With machine learning, high-throughput screening of REE-bearing coal ashes can be fulfilled based on a simple measurement.
- 24 October 2022
- Published by Springer Nature in Research Square
Open AccessPredicting Rare Earth Elements Concentration in Coal Ashes with Multi-Task Neural Networks
The increasing demand for rare earth elements (REEs) makes them a scarce strategic resource for technical developments. In that regard, harvesting REEs from coal ashes—a waste byproduct from coal power plants—offers an alternative solution to conventional ore-based extraction. However, this approach is bottlenecked by our ability to screen coal ashes bearing large concentrations of REEs from feedstocks—since measuring the REE content in ashes is a time-consuming and costly task requiring advanced analytical tools. Here, we propose a machine learning approach to predict the REE contents based on the bulk composition of coal ashes (which is easily measurable under the current testing protocol). We introduce a multi-task neural network that simultaneously predicts the contents of different REEs and, importantly, exhibits notably improved accuracy than the single-task models. Further model analyses reveal key data patterns for screening coal ashes with high REE concentrations.Teaser: With machine learning, high-throughput screening of REE-bearing coal ashes can be fulfilled based on a simple measurement.
- 22 September 2022
- Published by ASTM International in Advances in Civil Engineering Materials
- Vol. 11 (2) , 587-602
Interpreting the Strength Activity Index of Fly Ash with Machine Learning
Fly ash from the coal combustion at electric plants is commonly used for partially replacing portland cement in concrete production. Because of the varying nature of the coal source and the different processing protocols, different fly ashes exhibit wide ranges of physical and chemical characteristics, resulting in distinct impacts on concrete strength. Thus far, the most adopted method for assessing a given fly ash is specified by ASTM C618, Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete, wherein a series of influential chemical and physical features can be correlated to fly ash’s strength activity index (SAI). However, limited knowledge is available on how exactly the individual material attribute affects SAI, so accurately predicting the SAI remains out of reach. Here, we take advantage of recent advances in machine learning to reveal the origins of fly ash’s SAI. Leveraging a data set comprising 2,158 fly ash samples, we trained neural network models to predict 28-day SAI based on the sole knowledge of ASTM C618 material attributes. The results demonstrate that SAI is a complex property that does not systematically follow the conventional Class C/F classification. To gain a deeper insight into this matter, we further quantify the influence of each attribute on SAI as captured by the machine learning model.
- 22 September 2022
- Published by ASTM International in Advances in Civil Engineering Materials
- Vol. 11 (2) , 587-602
Interpreting the Strength Activity Index of Fly Ash with Machine Learning
Fly ash from the coal combustion at electric plants is commonly used for partially replacing portland cement in concrete production. Because of the varying nature of the coal source and the different processing protocols, different fly ashes exhibit wide ranges of physical and chemical characteristics, resulting in distinct impacts on concrete strength. Thus far, the most adopted method for assessing a given fly ash is specified by ASTM C618, Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete, wherein a series of influential chemical and physical features can be correlated to fly ash’s strength activity index (SAI). However, limited knowledge is available on how exactly the individual material attribute affects SAI, so accurately predicting the SAI remains out of reach. Here, we take advantage of recent advances in machine learning to reveal the origins of fly ash’s SAI. Leveraging a data set comprising 2,158 fly ash samples, we trained neural network models to predict 28-day SAI based on the sole knowledge of ASTM C618 material attributes. The results demonstrate that SAI is a complex property that does not systematically follow the conventional Class C/F classification. To gain a deeper insight into this matter, we further quantify the influence of each attribute on SAI as captured by the machine learning model.
- Open Access
- 1 November 2021
- Published by Elsevier in Cement and Concrete Composites
- Vol. 124
Influence of fine inclusions on the morphology and mechanical performance of lightweight foam concrete
- 1 November 2021
- Published by Elsevier in Cement and Concrete Composites
- Vol. 124
Open AccessInfluence of fine inclusions on the morphology and mechanical performance of lightweight foam concrete
- Open Access
- 1 November 2021
- Published by Elsevier in Knowledge-Based Systems
- Vol. 231
EBOD: An ensemble-based outlier detection algorithm for noisy datasets
- 1 November 2021
- Published by Elsevier in Knowledge-Based Systems
- Vol. 231
Open AccessEBOD: An ensemble-based outlier detection algorithm for noisy datasets
- 7 May 2021
- Published by ASTM International in Journal of Testing and Evaluation
- Vol. 50 (1) , 522-533
Measuring Dynamic Young’s Modulus of Low-Density Foam Concrete Using Resonant Frequency Test
Foam concrete is a cellular material mainly used in construction. Because of the low-strength feature of foam concrete, conventional concrete testing methods such as strength measurement are insufficient to characterize the key material attributes of foam concrete, especially for the low-density foam that is commonly designed for functional purposes. Inspired by the nondestructive feature of resonant frequency test, this study investigates the feasibility of using this method to evaluate the dynamic Young’s modulus of foam concrete, especially at the low-density range. Three variables that typically alter foam concrete mechanical properties are considered for evaluating the robustness of this approach—bulk density, water-to-cementitious ratio, and fly ash replacement ratio. The fundamental transverse frequency is found to be most suitable for determining the foam modulus. The experimental results demonstrate good accuracy of using this approach for measuring the dynamic Young’s modulus of different foam concrete samples. It is also confirmed that the foam modulus can be accurately designed if the bulk density of this material and solid modulus of its base cement paste are known.
- 7 May 2021
- Published by ASTM International in Journal of Testing and Evaluation
- Vol. 50 (1) , 522-533
Measuring Dynamic Young’s Modulus of Low-Density Foam Concrete Using Resonant Frequency Test
Foam concrete is a cellular material mainly used in construction. Because of the low-strength feature of foam concrete, conventional concrete testing methods such as strength measurement are insufficient to characterize the key material attributes of foam concrete, especially for the low-density foam that is commonly designed for functional purposes. Inspired by the nondestructive feature of resonant frequency test, this study investigates the feasibility of using this method to evaluate the dynamic Young’s modulus of foam concrete, especially at the low-density range. Three variables that typically alter foam concrete mechanical properties are considered for evaluating the robustness of this approach—bulk density, water-to-cementitious ratio, and fly ash replacement ratio. The fundamental transverse frequency is found to be most suitable for determining the foam modulus. The experimental results demonstrate good accuracy of using this approach for measuring the dynamic Young’s modulus of different foam concrete samples. It is also confirmed that the foam modulus can be accurately designed if the bulk density of this material and solid modulus of its base cement paste are known.