Search

This study presents a diagnostic framework for healthcare using explainable artificial intelligence and machine learning. Previous studies show that the ability of prediction models greatly depends on the relevance and independence of the feature set. Hence, various feature selection methods are presented in the literature. Moreover, the previous medical diagnostic models using neural networks offer accurate predictions, however, without generating explicit decision rules. A novel research framework with incremental feature selection and interpretable machine learning is proposed. First non-redundant and relevant features are selected. Further, the initial weights obtained during feature learning are fed to the interpretable neural network to obtain global and local explanations. This proposed research framework is demonstrated with an open-source medical dataset related to glioma, and the best-fit model is obtained. Moreover, an app for glioma grading is developed with the underlying predictive model to offer decision support to physicians and patients. [ABSTRACT FROM AUTHOR]

With the growth of the manufacturing industry, any unexpected failure of industrial equipment or machine breakdown can result in severe financial loss for the business. This is why it is critical to have a strategy for early detection and prediction for the failures. Predictive Maintenance encompasses all operational techniques and actions necessary to maintain machine availability and prevent downtime. The purpose of this article is to implement the machine-learning algorithm to develop a predictive maintenance and condition-based maintenance (CBM) systems. A real-world application of existing machine learning techniques for predictive maintenance was proposed and implemented using an artificial neural network (ANN). The results indicated that all predicted failures were correctly classified, with an overall accuracy of 99.9 percent. [ABSTRACT FROM AUTHOR]

This study presents a diagnostic framework for healthcare using explainable artificial intelligence and machine learning. Previous studies show that the ability of prediction models greatly depends on the relevance and independence of the feature set. Hence, various feature selection methods are presented in the literature. Moreover, the previous medical diagnostic models using neural networks offer accurate predictions, however, without generating explicit decision rules. A novel research framework with incremental feature selection and interpretable machine learning is proposed. First non-redundant and relevant features are selected. Further, the initial weights obtained during feature learning are fed to the interpretable neural network to obtain global and local explanations. This proposed research framework is demonstrated with an open-source medical dataset related to glioma, and the best-fit model is obtained. Moreover, an app for glioma grading is developed with the underlying predictive model to offer decision support to physicians and patients. [ABSTRACT FROM AUTHOR]

With the growth of the manufacturing industry, any unexpected failure of industrial equipment or machine breakdown can result in severe financial loss for the business. This is why it is critical to have a strategy for early detection and prediction for the failures. Predictive Maintenance encompasses all operational techniques and actions necessary to maintain machine availability and prevent downtime. The purpose of this article is to implement the machine-learning algorithm to develop a predictive maintenance and condition-based maintenance (CBM) systems. A real-world application of existing machine learning techniques for predictive maintenance was proposed and implemented using an artificial neural network (ANN). The results indicated that all predicted failures were correctly classified, with an overall accuracy of 99.9 percent. [ABSTRACT FROM AUTHOR]

This study presents a diagnostic framework for healthcare using explainable artificial intelligence and machine learning. Previous studies show that the ability of prediction models greatly depends on the relevance and independence of the feature set. Hence, various feature selection methods are presented in the literature. Moreover, the previous medical diagnostic models using neural networks offer accurate predictions, however, without generating explicit decision rules. A novel research framework with incremental feature selection and interpretable machine learning is proposed. First non-redundant and relevant features are selected. Further, the initial weights obtained during feature learning are fed to the interpretable neural network to obtain global and local explanations. This proposed research framework is demonstrated with an open-source medical dataset related to glioma, and the best-fit model is obtained. Moreover, an app for glioma grading is developed with the underlying predictive model to offer decision support to physicians and patients. [ABSTRACT FROM AUTHOR]

With the growth of the manufacturing industry, any unexpected failure of industrial equipment or machine breakdown can result in severe financial loss for the business. This is why it is critical to have a strategy for early detection and prediction for the failures. Predictive Maintenance encompasses all operational techniques and actions necessary to maintain machine availability and prevent downtime. The purpose of this article is to implement the machine-learning algorithm to develop a predictive maintenance and condition-based maintenance (CBM) systems. A real-world application of existing machine learning techniques for predictive maintenance was proposed and implemented using an artificial neural network (ANN). The results indicated that all predicted failures were correctly classified, with an overall accuracy of 99.9 percent. [ABSTRACT FROM AUTHOR]

Technological advancements in Machine learning, artificial intelligence (AI), and data science are bringing industries to the era of the fourth industrial revolution. The application of machine learning in chemical engineering is in the domains of process modeling, optimization, and predictive analysis. Traditional process modeling relies heavily on first-principal methods, which, while accurate, are computationally demanding and are non-flexible for variable process conditions. Green methanol produced through the power-to-liquid (PtL) process has gained significant popularity due to its various applications in household items, as a raw material for manufacturing valuable chemicals, and as a fuel both in blend or pure form. In today's competitive and uncertain chemical industry market, fast and accurate models are required to predict the plant output. This work aims to develop a surrogate model of the methanol production process based on the data-driven technique and using machine learning to predict energy requirements, final product purity, and methanol production rate. The effect of the sampling size and sampling technique (mainly Latin-Hypercube Sampling - LHS, Monte Carlo, and SOBOL) on the performance of the surrogate model is evaluated. A comparative analysis of different machine learning (e.g., XG-Boost, Random Forest, Decision Tree, Support Vector Regression) and Deep learning models (e.g., Artificial Neural Networks) is conducted using metrics such as coefficient of determination (R²), mean-squared error (MSE), and mean-absolute-error (MAE). Additionally, this work explores the use of these trained machine learning models in optimizing process conditions to maximize production rate, enhance product purity, and reduce energy requirements. [ABSTRACT FROM AUTHOR]

Hydrogen Sulphide Methane Reformation (HSMR) is a viable alternative for simultaneous H2S valorization and hydrogen production, offering a carbon-neutral alternative. Concurrently, the industrial revolution, driven by emerging technologies like machine learning and artificial intelligence is reshaping chemical industry as well. machine learning offers robust models for accurate process forecasting, overcoming computational demands and inflexibility. This work focuses on developing a surrogate model for the HSMR process, a promising, waste to energy, route for sustainable hydrogen production. Integrating machine learning with design of experiment techniques, the study systematically generates data using Aspen plus v11 simulation and employs established adaptive space filling sampling techniques. Models such as Linear Regression, Support Vector Regression, Random Forest Regression, Extreme Gradient Boosting, and Artificial Neural Networks was trained on the refined data set, an accuracy of up to 97% in predicting the process outputs was achieved. Furthermore, this research extends to the optimization of an artificial neural network (ANN) surrogate model using evolutionary algorithms, specifically genetic algorithms and differential evolution. within the Python environment, this optimization aims to identify optimal process conditions for HSMR. A comparative analysis with the base case provides insights into the effectiveness of these evolutionary algorithms in enhancing the performance of the ANN model. This work contributes to enhancing the efficiency and competitiveness of HSMR, bridging the gap between machine learning capabilities and sustainable energy solutions. [ABSTRACT FROM AUTHOR]

Hydrogen Sulphide Methane Reformation (HSMR) is a viable alternative for simultaneous H2S valorization and hydrogen production, offering a carbon-neutral alternative. Concurrently, the industrial revolution, driven by emerging technologies like machine learning and artificial intelligence is reshaping chemical industry as well. machine learning offers robust models for accurate process forecasting, overcoming computational demands and inflexibility. This work focuses on developing a surrogate model for the HSMR process, a promising, waste to energy, route for sustainable hydrogen production. Integrating machine learning with design of experiment techniques, the study systematically generates data using Aspen plus v11 simulation and employs established adaptive space filling sampling techniques. Models such as Linear Regression, Support Vector Regression, Random Forest Regression, Extreme Gradient Boosting, and Artificial Neural Networks was trained on the refined data set, an accuracy of up to 97% in predicting the process outputs was achieved. Furthermore, thisresearch extends to the optimization of an artificial neural network (ANN) surrogate model using evolutionary algorithms, specifically genetic algorithms and differential evolution. within the Python environment, this optimization aims to identify optimal process conditions for HSMR. A comparative analysis with the base case provides insights into the effectiveness of these evolutionary algorithms in enhancing the performance of the ANN model. This work contributes to enhancing the efficiency and competitiveness of HSMR, bridging the gap between machine learning capabilities and sustainable energy solutions. [ABSTRACT FROM AUTHOR]

Technological advancements in Machine learning, artificial intelligence (AI), and data science are bringing industries to the era of the fourth industrial revolution. The application of machine learning in chemical engineering is in the domains of process modeling, optimization, and predictive analysis. Traditional process modeling relies heavily on first-principal methods, which, while accurate, are computationally demanding and are non-flexible for variable process conditions. Green methanol produced through the power-to-liquid (PtL) process has gained significant popularity due to its various applications in household items, as a raw material for manufacturing valuable chemicals, and as a fuel both in blend or pure form. In today's competitive and uncertain chemical industry market, fast and accurate models are required to predict the plant output. This work aims to develop a surrogate model of the methanol production process based on the data-driven technique and using machine learning to predict energy requirements, final product purity, and methanol production rate. The effect of the sampling size and sampling technique (mainly Latin-Hypercube Sampling - LHS, Monte Carlo, and SOBOL) on the performance of the surrogate model is evaluated. A comparative analysis of different machine learning (e.g., XG-Boost, Random Forest, Decision Tree, Support Vector Regression) and Deep learning models (e.g., Artificial Neural Networks) is conducted using metrics such as coefficient of determination (R²), mean-squared error (MSE), and mean-absolute-error (MAE). Additionally, this work explores the use of these trained machine learning models in optimizing process conditions to maximize production rate, enhance product purity, and reduce energy requirements. [ABSTRACT FROM AUTHOR]

Technological advancements in Machine learning, artificial intelligence (AI), and data science are bringing industries to the era of the fourth industrial revolution. The application of machine learning in chemical engineering is in the domains of process modeling, optimization, and predictive analysis. Traditional process modeling relies heavily on first-principal methods, which, while accurate, are computationally demanding and are non-flexible for variable process conditions. Green methanol produced through the power-to-liquid (PtL) process has gained significant popularity due to its various applications in household items, as a raw material for manufacturing valuable chemicals, and as a fuel both in blend or pure form. In today's competitive and uncertain chemical industry market, fast and accurate models are required to predict the plant output. This work aims to develop a surrogate model of the methanol production process based on the data-driven technique and using machine learning to predict energy requirements, final product purity, and methanol production rate. The effect of the sampling size and sampling technique (mainly Latin-Hypercube Sampling - LHS, Monte Carlo, and SOBOL) on the performance of the surrogate model is evaluated. A comparative analysis of different machine learning (e.g., XG-Boost, Random Forest, Decision Tree, Support Vector Regression) and Deep learning models (e.g., Artificial Neural Networks) is conducted using metrics such as coefficient of determination (R²), mean-squared error (MSE), and mean-absolute-error (MAE). Additionally, this work explores the use of these trained machine learning models in optimizing process conditions to maximize production rate, enhance product purity, and reduce energy requirements. [ABSTRACT FROM AUTHOR]

Hydrogen Sulphide Methane Reformation (HSMR) is a viable alternative for simultaneous H2S valorization and hydrogen production, offering a carbon-neutral alternative. Concurrently, the industrial revolution, driven by emerging technologies like machine learning and artificial intelligence is reshaping chemical industry as well. machine learning offers robust models for accurate process forecasting, overcoming computational demands and inflexibility. This work focuses on developing a surrogate model for the HSMR process, a promising, waste to energy, route for sustainable hydrogen production. Integrating machine learning with design of experiment techniques, the study systematically generates data using Aspen plus v11 simulation and employs established adaptive space filling sampling techniques. Models such as Linear Regression, Support Vector Regression, Random Forest Regression, Extreme Gradient Boosting, and Artificial Neural Networks was trained on the refined data set, an accuracy of up to 97% in predicting the process outputs was achieved. Furthermore, this research extends to the optimization of an artificial neural network (ANN) surrogate model using evolutionary algorithms, specifically genetic algorithms and differential evolution. within the Python environment, this optimization aims to identify optimal process conditions for HSMR. A comparative analysis with the base case provides insights into the effectiveness of these evolutionary algorithms in enhancing the performance of the ANN model. This work contributes to enhancing the efficiency and competitiveness of HSMR, bridging the gap between machine learning capabilities and sustainable energy solutions. [ABSTRACT FROM AUTHOR]