Scopus ID: 85159477384
2022 IEEE 14th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), (2022), pp. 1-6
Conference Paper | Published: January 1, 2022
Abstract
Carbonation is a dangerous threat to concrete since it reduces the alkalinity of normal or self-compacting concrete (SCC), allowing iron to corrode and spall the cover. The goal of this research is to use an artificial neural network to create a corrosion prediction model for steel in self-compacting concrete that has been subjected to carbonation. In this study, MATLABR2019a was used to create a feedforward back propagation neural network. As a training function, the researchers utilized the Levenberg-Marquardt back propagation (TRAINLM) which adjusts weights and bias values using Levenberg-Marquardt optimization. The researchers used gradient descent with momentum weight/bias learning (LEARNGDM) for the adaptation learning function, which is a technique that aids the gradient in determining which way to go. The network’s performance was measured using the mean square error (MSE). The Hyperbolic tangent sigmoid transfer function (TANSIG) was also employed as the transfer function since the values obtained by this function range from +1 to -1, considering both the positive and negative aspects of the parameter. To minimize overfitting, the number of hidden nodes should be fewer than the number of input parameters. The researchers tested 4-12 hidden nodes. Modeling was done using data from 102 experimental studies of self-compacting concrete exposed to corrosion. Using feed-forward back propagation ANN with 1 hidden layer and 8 hidden nodes, a Pearson R-value of 0.98748 and a mean square error of 0.5725 were obtained. The factor that most affect the carbonation depth were water-cement ratio and fly ash content. The suggested model was able to analytically describe the connection and behaviors of the various mixtures to the carbonation depth in the parametric investigation. The parameters characteristics were likewise described by the model.
Scopus ID: 85159404536
2022 IEEE 14th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), (2022), pp. 1-5
Conference Paper | Published: January 1, 2022
Abstract
Technology in construction has been attempting to discover eco-friendly materials, to identify waste products that can be processed as an alternative supplement to cement. The study utilizes sugar cane bagasse ash (SCBA) to produce a bagasse ash blended concrete. There were 15 previous experimental studies that were gathered and analyzed which have the same variables in terms of determining the mechanical proeprties of the blended concrete with SCBA. The variables that were considered are: compressive strength, cement content (CC), fine aggregate (FA) and coarse aggregate (CA) content, water-cement ratio (W/C), water content (WC), and sugarcane bagasse ash content. 74 different data sets in all were obtained. The study's goal is to develop a prediction model for estimating the mechanical properties of concrete made with SCBA. The work employed MATLAB R2021a neural network (NN) toolbox for model development and simulation of the dataset with the use of the backpropagation ANN. The best model was observed to have a structure of 7-7-1 (input-hidden-output) having the highest R all value and lowest AIC value, with a mean absolute percentage error (MAPE) is 3.718% considered to be a highly accurate model. The relative importance (RI) showed that the FA, CA, and CC were the most significant factors to the CS while water and SCBA were the least influential parameters. The overall findings reveal that the MAPE of the compression strength prediction model decreased from 3.718% to 3.519% exhibiting a 5.35% improvement in the model’s performance.
Scopus ID: 85127588959
2021 IEEE 13th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), (2021), pp. 1-6
John Mark Cagurungan, Royvin Factuar, ... Jon Arnel S. Telan
Conference Paper | Published: January 1, 2021
Abstract
Solid waste generation is one of the world’s most prevalent challenges, especially in places with crowded populations and inadequate solid waste disposal strategies. There are several extant influencing variables on solid waste creation. In this regard, the researchers focus on five (5) elements or categories that contributed the most to solid trash generation. The researchers sought to determine which one has the greatest influence on solid waste generation in Barangay Sagrada Familia among these five categories. This will contribute to their future solid waste management plan through minimizing, segregating, and recycling the solid waste, which is one of the causes of their flooding problem. ANN (Artificial Neural Network) is a simplified computational brain model that is one of the most often utilized artificial intelligence in solid waste management. To get the desired outcomes, Matrix Laboratory (MATLAB) testing is essential. The researchers gathered information from studies, theories, and literature in the field. The researchers then performed a survey to gather data and existing data in the barangay and used Excel and Matrix Laboratory (MATLAB) to construct the model for a Neural Network analysis. Finally, the authors analyzed the Neural Network, with the goal value varying according to Pearson’s Correlation Coefficient (R).
Scopus ID: 85127585497
2021 IEEE 13th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), (2021), pp. 1-6
Sophia Chloe Caress, Angela Abigail Belen, ... Melvin B. Solomon
Conference Paper | Published: January 1, 2021
Abstract
Agriculture relies heavily on weather forecasts, and a reliable weather forecasting system can help mitigate the calamities which can affect this industry. Rainfall and meteorological drought duration forecasting are some of the most important yet challenging tasks. This paper presents the creation of feedforward backpropagation artificial neural networks for daily rainfall forecasting and monthly meteorological drought forecasting. Artificial Neural Networks can capture the variability of these phenomena. Rainfall data from nine stations all over Albay, the Philippines, spanning from 1967 to 2000, were used to create the models. The input parameters used for developing the models for daily rainfall forecasting were 14-day antecedent rainfall, current-day rainfall, relative humidity, mean temperature, and sunshine duration. The monthly meteorological drought forecasting parameters were 1-month SPI, current-month rainfall, relative humidity, mean temperature, and sunshine duration. Having the results presented in this paper, the performance of the ANN Models of the stations were compared based on R and RMSE. The rainfall forecasting models and meteorological drought forecasting models have provided satisfactory performance. A satisfactory performance for forecasting has an R-value ranging from 0.2 to 0.5. Sensitivity analysis indicated that the most significant parameter for rainfall forecast is the relative humidity and mean temperature for drought forecast.
Scopus ID: 85127555504
2021 IEEE 13th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), (2021)
Conference Paper | Published: January 1, 2021
Abstract
In a worldwide perspective of the most common cancer diseases, cervical cancer is ranked fourth most frequent whereas the worldwide mortality rate is at 54.56%. In the Philippines, the second leading site among women is cervical cancer next to breast cancer. Research shows that cervical cancer is one of the most treatable cancer forms if detected and managed early. Currently, the most reliable diagnosis and prevention method of cervical cancer is thru a regular testing via Pap Smear test and HPV vaccination being performed in hospitals worldwide. However, according to the Centers for Disease Control and Prevention in California, the cervical cancer screening rate of regular testing in hospitals went down significantly during the stay-at-home order by the government due to the COVID-19 pandemic. Also, there are limited research based on the behavior information in relation to cervical cancer risk prediction, but existing studies proves the possibility of the risk prediction based on behavior information. This paper presents an Artificial Neural Network-based model for early cervical cancer risk detection based on behavior information. The neural network was trained using scaled conjugate gradient back propagation. The system showed 98% overall correctness in early cervical cancer risk prediction.
Scopus ID: 85105008189
2020 IEEE 12th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), (2020), pp. 1-5
Conference Paper | Published: December 3, 2020
Abstract
Diabetes is a disease which chronic in nature, which is caused by an elevated blood sugar (or blood glucose) level. The metabolic disease is linked to several potential serious organ complications including nerves, kidneys, eyes, blood vessels, and the heart. According to the International Diabetes Federation, in 2019, about 2 million deaths were recorded worldwide due to diabetes. Furthermore, according to Philippine Statistics Authority (PSA), Diabetes Mellitus is considered as the fifth main cause of in the Philippines in the past years and in a 2015 study, about 1.7 million Filipinos are still undiagnosed of diabetes. Therefore, several machine learning-based techniques were developed for diabetes risk prediction. However, these works have yet to utilize artificial neural networks using the symptom information of suspected diabetic patients. This research paper demonstrated an ANN-based diabetes risk classification based on the symptom information of patients. The scaled conjugate gradient backpropagation technique was utilized for neural network training process. The classification system showed 99.2% overall correctness in determining the likelihood of diabetes.
Scopus ID: 85087445229
Clean Technologies and Environmental Policy, (2020), Vol. 22, No. 7, pp. 1449-1465
Ivan Henderson V. Gue, Aristotle T. Ubando, ... Raymond R. Tan
Journal Article | Published: September 1, 2020
Abstract
Computational and statistical tools help manage the prevailing challenges of the 17 Sustainable Development Goals (SDGs) by providing meticulous understanding of contemporary issues. However, complex challenges are difficult to handle with conventional techniques, resulting to the need for more advanced methods. Artificial neural networks (ANNs) are often used as an advanced approach in modelling complex behaviour of systems. Evaluating the current utilization of ANNs helps researchers gauge their applicability to SDG-related issues. The gaps among the studied SDGs need to be addressed through a comprehensive survey of the state-of-the-art literature. Hence, this work reviews published journal articles on the application of ANNs in resolving issues of the SDGs. This review identifies the current trends and limitations of ANN for SDG, and discusses its prominent applications and field of utilization. Descriptive and content analysis of journal articles is performed for this review. Journal articles from the Scopus database reveal Clean Water and Sanitation, Affordable and Clean Energy, Sustainable Cities and Communities, and Responsible Consumption and Production are the most popular subject matter for modelling and forecasting. New innovative functions include feature selection, kriging, and simulation. The main contribution of this work is a comprehensive mapping of the current state of this area of research. This work aims to aid future researchers to recognize further possible uses of ANNs with respect to the SDGs.
Scopus ID: 85079735115
International Journal of GEOMATE, (2020), Vol. 18, No. 65, pp. 179-184
Nolan C. Concha
& Andres Winston OretaJournal Article | Published: January 1, 2020
Abstract
The composite action of reinforcement in the surrounding concrete involve a complex and non-linear mechanism.Inadequate understanding of the underlying interactions may lead to designs with insufficient amount of bond resistance of reinforcing bars in concrete structures.To investigate the effects of various parameters on the bond strength of steel bars in concrete, 54 cube samples with varying embedded reinforcements and strengths were prepared. The samples were cured for 28 days and tested using ultrasonic pulse velocity (UPV) test for sample homogeneity and single pull out test for bond strength.Data gathered in the experiment were used in the development of bond strength model as a function of compressive strength, concrete cover to rebar diameter ratio, embedment length, and UPV using artificial neural network (ANN). Of all the bond strength models considered from various literatures, the neural network model provided the most satisfactory prediction results in good agreement with the bond strength values obtained from the experiment. The UPV parameter was found to be one of the most significant predictors in the neural network model having a relative importance of 20.57%. This suggest that the robust prediction performance of the bond model was attributed to this essential component of the model. The proposed model of this study can be used as baseline information and rapid non-destructive assessment for zone wise strengthening in reinforced concrete.
Scopus ID: 85084746978
2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management ( HNICEM ), (2019), pp. 1-5
Conference Paper | Published: November 1, 2019
Abstract
Shrimp farming is a highly profitable business in the aquaculture industry. The farming profitability can be achieved by the implementation of better management practices in conjunction with optimal shrimp feed management and growth monitoring. Manual measurement for shrimp growth on a large population is a tedious and difficult task. Underfeeding results to lower growth rate, and overfeeding results to environmental pollution. Automated, continuous, and non-invasive methods therefore such as computer vision are being increasingly employed. However, existing researches of vision-based measurement of growth parameters are not yet incorporated to shrimp feed management. This paper presented an Artificial Neural Network-based decision support system of classifying feed type whether starter, grower or finisher using area, length and weight derived from image processing techniques. The neural network was trained using scaled conjugate gradient back propagation. The decision support system exhibited promising results in feed type classification.
Scopus ID: 85084763596
2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management ( HNICEM ), (2019), pp. 1-4
Conference Paper | Published: November 1, 2019
Abstract
Cost estimation of building demolition, like any engineering project, requires ample amount of time and experience to accomplish since it involves calculations of complex relationships between its influencing factors. Since artificial neural networks (ANNs) are known to be effective in the cost-forecasting domain with complex parameters involve, the study aims to develop an ANN that can predict building demolition cost in Quezon City. One-hundred demolition projects from the Department of Building Official in Quezon City were gathered, evaluated and divided randomly into two sets: 90% for training, validation and internal testing and 10% for external application. Nine demolition cost-influencing factors were identified, namely: building condition, materials and classification, number of floors, total floor area, site accessibility, location, demolition methods used and debris removal options. The training was applied with feedforward backpropagation algorithm. The resulting architecture for the selected ANN model consists of 12 hidden nodes. The model tested and was successful in predicting demolition cost in Quezon City with an average accuracy rating of 90.21%.