\"Complex materials consisting of multiple chemical components/phases have been widely used in different engineering applications that have strict criteria. To optimize the mixture design of complex materials, researchers invest enormous resources in cumbersome experiments. However, due to substantial variations in precursors and processing techniques, it has been challenging to develop theoretical frameworks (i.e., empirical models) that could produce reliable predictions of properties of complex materials. In recent years, scientists have harnessed the power of machine learning (ML) and “Big” data to uncover the underlying mixture design-property correlations and produce high-fidelity predictions of properties of complex materials. The ML models autonomously learn cause-effect correlations from the training dataset, and then capitalize on such knowledge to produce predictions on a new data domain. This research consists of six studies. The first three studies focus on utilizing ML models to predict and optimize the fresh and hardened properties of Portland cement (PC) at different ages. The fourth study presents the ML models to predict the compressive strength of alkali-activated cement – a sustainable cement can replace PC – in relation to topological parameters and mixture designs. In the fifth paper, six ML models are employed to predict and optimize the dissolution kinetics of bioactive glasses. The sixth study compares the performance of analytical and ML models in predicting the sulfur solubility of nuclear waste glasses. In some aforesaid studies, closed-form analytical models are developed to predict material properties based on outcomes from ML models\"--Abstract, p. iv
","abstract_html":"<p>"Complex materials consisting of multiple chemical components/phases have been widely used in different engineering applications that have strict criteria. To optimize the mixture design of complex materials, researchers invest enormous resources in cumbersome experiments. However, due to substantial variations in precursors and processing techniques, it has been challenging to develop theoretical frameworks (i.e., empirical models) that could produce reliable predictions of properties of complex materials. In recent years, scientists have harnessed the power of machine learning (ML) and “Big” data to uncover the underlying mixture design-property correlations and produce high-fidelity predictions of properties of complex materials. The ML models autonomously learn cause-effect correlations from the training dataset, and then capitalize on such knowledge to produce predictions on a new data domain. This research consists of six studies. The first three studies focus on utilizing ML models to predict and optimize the fresh and hardened properties of Portland cement (PC) at different ages. The fourth study presents the ML models to predict the compressive strength of alkali-activated cement – a sustainable cement can replace PC – in relation to topological parameters and mixture designs. In the fifth paper, six ML models are employed to predict and optimize the dissolution kinetics of bioactive glasses. The sixth study compares the performance of analytical and ML models in predicting the sulfur solubility of nuclear waste glasses. In some aforesaid studies, closed-form analytical models are developed to predict material properties based on outcomes from ML models"--Abstract, p. iv</p>","abstract_has_math":false,"creators":["Han, Taihao"],"institution":"Missouri University of Science and Technology","degree_name":"Ph. D. in Materials Science and Engineering","degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-07-24T03:18:18Z","subjects":["Materials Science and Engineering"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://scholarsmine.mst.edu/doctoral_dissertations/3231","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Han, Taihao"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:type","label":"Dc Type","values":["Dissertation - Open Access"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph. D. in Materials Science and Engineering"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["Missouri University of Science and Technology"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Materials Science and Engineering"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://scholarsmine.mst.edu/doctoral_dissertations/3231"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["\"Complex materials consisting of multiple chemical components/phases have been widely used in different engineering applications that have strict criteria. To optimize the mixture design of complex materials, researchers invest enormous resources in cumbersome experiments. However, due to substantial variations in precursors and processing techniques, it has been challenging to develop theoretical frameworks (i.e., empirical models) that could produce reliable predictions of properties of complex materials. In recent years, scientists have harnessed the power of machine learning (ML) and “Big” data to uncover the underlying mixture design-property correlations and produce high-fidelity predictions of properties of complex materials. The ML models autonomously learn cause-effect correlations from the training dataset, and then capitalize on such knowledge to produce predictions on a new data domain. This research consists of six studies. The first three studies focus on utilizing ML models to predict and optimize the fresh and hardened properties of Portland cement (PC) at different ages. The fourth study presents the ML models to predict the compressive strength of alkali-activated cement – a sustainable cement can replace PC – in relation to topological parameters and mixture designs. In the fifth paper, six ML models are employed to predict and optimize the dissolution kinetics of bioactive glasses. The sixth study compares the performance of analytical and ML models in predicting the sulfur solubility of nuclear waste glasses. In some aforesaid studies, closed-form analytical models are developed to predict material properties based on outcomes from ML models\"--Abstract, p. iv
"]},{"key":"dc:title","label":"Title","values":["DEVELOPMENT AND APPLICATION OF ENSEMBLE MACHINE LEARNING ALGORITHMS TO ENABLE A PRIORI, HIGH-FIDELITY, AND PROMPT PREDICTIONS AND OPTIMIZATIONS OF COMPLEX MATERIALS AND SYSTEMS"]}]}],"canonical_facts":{"dc:creator":["Han, Taihao"],"dc:description.abstract":["\"Complex materials consisting of multiple chemical components/phases have been widely used in different engineering applications that have strict criteria. To optimize the mixture design of complex materials, researchers invest enormous resources in cumbersome experiments. However, due to substantial variations in precursors and processing techniques, it has been challenging to develop theoretical frameworks (i.e., empirical models) that could produce reliable predictions of properties of complex materials. In recent years, scientists have harnessed the power of machine learning (ML) and “Big” data to uncover the underlying mixture design-property correlations and produce high-fidelity predictions of properties of complex materials. The ML models autonomously learn cause-effect correlations from the training dataset, and then capitalize on such knowledge to produce predictions on a new data domain. This research consists of six studies. The first three studies focus on utilizing ML models to predict and optimize the fresh and hardened properties of Portland cement (PC) at different ages. The fourth study presents the ML models to predict the compressive strength of alkali-activated cement – a sustainable cement can replace PC – in relation to topological parameters and mixture designs. In the fifth paper, six ML models are employed to predict and optimize the dissolution kinetics of bioactive glasses. The sixth study compares the performance of analytical and ML models in predicting the sulfur solubility of nuclear waste glasses. In some aforesaid studies, closed-form analytical models are developed to predict material properties based on outcomes from ML models\"--Abstract, p. iv
"],"dc:identifier":["https://scholarsmine.mst.edu/doctoral_dissertations/3231"],"dc:subject":["Materials Science and Engineering"],"dc:title":["DEVELOPMENT AND APPLICATION OF ENSEMBLE MACHINE LEARNING ALGORITHMS TO ENABLE A PRIORI, HIGH-FIDELITY, AND PROMPT PREDICTIONS AND OPTIMIZATIONS OF COMPLEX MATERIALS AND SYSTEMS"],"dc:type":["Dissertation - Open Access"],"thesis:degree_name":["Ph. D. in Materials Science and Engineering"],"thesis:institution_name":["Missouri University of Science and Technology"]},"updated_at":"2026-07-24T03:18:18Z"}