Pocholo James M. Loresco
29 Publications
Scopus ID: 85159483095
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
By early 2020, COVID-19 has caused a global pandemic which led to an enormous number of challenges worldwide in various sectors. The Philippine government has implemented multiple quarantine guidelines and travel restrictions to ensure the people’s health and safety. However, the International Labour Organization projected an initial economic and labor market disruption affecting 11 million workers, or about 25% of the Philippine workforce, due to the pandemic. Therefore, the government, thru the concerned agencies continues to encourage employers to implement alternative work plans such as a work-from-home (WFH) operation in compliance with the established regulations in line with existing laws and policies. In line with the telecommuting concept, various research has already been performed, however, some were regarded inconclusive and require further study. Hence, in this study, a Web application was developed along with an embedded fuzzy model to evaluate the telecommuting capability assessment of employees. The proposed web application with embedded fuzzy model is capable of providing capability assessment using the four main input variables which are also relatively characterized for possible telecommuting cost assessment.
Scopus ID: 85159471855
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
Structural health monitoring (SHM) was developed to provide diagnosis of the state of civil structures, such as bridge and building, throughout their lifespan. The superstructure degrades over time mainly because of the continuous effect of environmental factors including temperature, wind, humidity, and traffic loading. Consequently, this study is concerned in reducing the masking effect of environmental factors, specifically on temperature. Principal component analysis (PCA), a supervised machine learning algorithm, was employed as the embedded statistical treatment for multidimensional reduction of feature matrix data to eliminate the temperature effect. Kaiser’s criterion eliminated data variance of almost 20% that may result to poor data reconstruction. Cumulative proportion (CP) criterion eliminated data variance around 4%, which is a better choice for deciding the number of principal components to eliminate. Thus, the proposed experimental study addressed successful temperature compensation from reinforced concrete bridge vibration data by using PCA and CP criterion.
Scopus ID: 85159411512
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
Determination of relay pairs from breakpoints in a given network is essential for maintaining the current protection system. Pairs of relays as primary or backup maintain the operation of the protection scheme within its zone of protection in tandem. All calculations and assumptions that are made in protection systems are based on breakpoints. It is inadequately documented that machine learning can be used to determine breakpoint sets and relay pairs. This paper presents the implementation of supervised decision tree machine learning approach for determining directional overcurrent relay breakpoint set in 3-bus networks. Using the one-hot encoding method, 45 input features are extracted from a matrix derived from 3-bus, 5-line network data. Bayesian optimization is used to further optimize the hyperparameters of each model for each of the break point set outputs. Tree diagrams are also provided here to assist in the interpretation of the decision rule resulting from the regression analysis. Experiment tests indicated that the proposed method shows promising results in determining breakpoint set in terms of RMSE.
Scopus ID: 85159479494
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
Pork meat is a very important source of proteins and other nutrients, so it requires a high level of quality. There is a serious health risk associated with the consumption of spoiled or contaminated pork meat, which is why it is extremely important to monitor its freshness. In this study, sensor arrays consisting of RGB IR sensors, thermal sensors, electronic noses (gas sensors) for detecting the color, temperature, and carbon dioxide and ammonia level of the pork meat were used to evaluate pork meat quality and estimate shelf life. The use of various supervised machine learning approaches has been applied with optimization to perform classification as to whether the meat was fresh or not, as well as regression analysis to predict the amount of exposure time for the meat that can be used in computing shelf-life estimates. Several high-performance algorithms were then tested, evaluated, and compared after hyperparameters of each model were optimized using grid search. As a result of a comparative analysis of the machine learning used, gentle boost ensembles outperformed other machine learning methods in detecting pork meat quality with 92.8% accuracy, while gaussian process regression predicted shelf life with the lowest RMSE, MSE and MAE.
Scopus ID: 85159218924
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
The Virtual Dietitian (VD) application is a nutrition knowledge-based system that generates personalized meal plans in accordance with the one-size-does-not-fit-all concept of precision nutrition. A subset of the population that was not involved in its four-part developmental study was gym and fitness enthusiasts despite them being important target users. As part of our quality improvement (QI) plan, we initiated a two-phase user testing to inform modifications to VD. We recruited a total of 30 users with prior experience in nutrition applications. In phase 1, they used the current version of VD for a week and answered a mixed-form questionnaire afterward. We used the same questionnaire from our previous study, which is composed of System Usability Scale (SUS) items and open-ended questions. After months of system modification, the same set of users evaluated again the new VD version after another week of use. A paired-sample t-test showed a statistically significant difference in SUS scores before (SUS = 79) and after (SUS = 82) modifying VD based on the suggestions of the participants (p = 0.005). Some new features include water tracker and reminder modules, Google Fit integration, and other nutrition support services (e.g., teleconsultation with registered dietitians). Although further refinements to VD are still needed, we were able to incorporate a QI initiative typically employed by healthcare organizations into software development for a better and improved personalized nutrition application.
Scopus ID: 85159453987
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
Tropical cyclones are a common occurrence in the Philippines since it is located in the Western North Pacific region. It is essential to monitor and measure rainfall as it has implications for reducing disaster risk, agricultural planning, and transportation planning. In this study, machine learning artificial neural networks are employed in the classification of acoustic data collected from a rain acoustic sensor (RAS) developed. Mel-frequency cepstrum coefficients are extracted and modified into inputs for an artificial neural network (ANN) that classifies rainfall into Light, Moderate, Heavy, Intense and Torrential according to the Philippine Atmospheric, Geophysical, and Astronomical Services Administration (PAGASA) Rainfall Classification System. Using scaled conjugate gradient backpropagation, the neural network was trained, and its results were compared with the standard tipping bucket rain gauge data. A 94.6% accuracy rate was achieved using five-fold cross-validation for classifying PAGASA’s rainfall data in the experimental study.
Scopus ID: 85159479828
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
The COVID-19 pandemic has caused disruption to the economy due to the increasing infection that affects the workforce in different sectors. The Philippine government has imposed lockdowns to control the spread of infection. This urged the different sectors to implement flexible work schedules or work from home setup. A work-from-home (WFH) setup burdens both the employee and employer by installing different equipment set-ups such as WiFi-equipped laptops, computers, tablets, or smartphones. However, the internet stability in some of the areas in the Philippines is not yet reliable. In this study, an application is used collect survey information and provide an estimate of the telework internet cost requirement of a given government employee or a given government employee implementing a work-from-home set up in their respective household. This involves survey results from different respondents who are currently on a work-from-home setup and significant factors from the survey have been analyzed using machine learning (ML) algorithms. Among the machine learning algorithms used, the ensemble bagged trees model outperformed the other ML models. This work can be extended by incorporating a wider scope of datasets from different industry doing work from home set-up. In addition, in terms of education, it is also recommended to determine the WFH set up not just with the government employee and employer but to also extend this into the education side.
Scopus ID: 85159482541
2022 IEEE 14th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), (2022), pp. 1-6
Ryan Rhay P. Vicerra, Rex Paolo C. Gamara, ... Andres Philip Mayol
Conference Paper | Published: January 1, 2022
Abstract
By the beginning of 2020, the illness had been named as COVID-19, which had spread due to its extreme severity affecting multiple industries and sectors throughout the world. To protect the public's health and safety, the Philippine government has established a number of quarantine regulations and travel restrictions in reaction to the current COVID-19 outbreak. Nonetheless, the ILO predicted that the pandemic would initially disrupt the economy and labor markets, affecting 11 million employees, or around 25% of the workforce in the Philippines. Therefore, the government continues to urge employers of local companies and enterprises to use alternative work plans, such as a WFH – work-from-home operation in accordance with the established policies. In line with the concept of telework, several studies have already been carried out, though some were declared inconclusive and require additional study. Hence, in this research, a mobile application was created to evaluate the employee’s telework capability assessment using a Fuzzy-based model which utilizes Google AppSheet, Apps Script, and Sheets. The developed mobile application is able to provide capacity evaluation utilizing the four key input variables, which are also reasonably characterized for potential telecommuting cost evaluation.
Scopus ID: 85112824462
International Journal of Emerging Technologies in Learning (iJET), (2021), Vol. 16, No. 15, pp. 31
Journal Article | Published: January 1, 2021
Abstract
Attaining high retention rates among engineering institutions is a predominant is-sue. A significant portion of engineering students face challenges of retention. Academic advising was implemented to resolve the issue. Decision support sys-tems were developed to support the endeavor. Machine learning have been inte-grated among such systems in predicting student performance accurately. Most works, however, rely on a black box model approach. Rule induction generates simpler if-then rules, exhibiting clearer understanding. As most research works considered attributes for positive academic performance, there is the need to con-sider ‘negative’ attributes. ‘Negative’ attributes are critical indicators to possibility of failure. This work applied rule induction techniques for course grade predic-tion using ‘negative’ attributes. The dataset is the academic performance of 48 mechanical engineering students taking a machine design course. Students’ at-tributes on workload, course repetition, and incurred absences are the predictors. This work implemented two rule induction techniques, rough set theory (RST) and adaptive neuro fuzzy inference system (FIS). Both models attained a classifi-cation accuracy of 70.83% with better performance for course grades of ‘Pass’ and ‘High’. RST generated 16 crisp rules while ANFIS generated 27 fuzzy rules, yielding significant insights. Results of this study can be used for comparative analysis of student traits between institutions. The illustrated framework can be used in formulating linguistic rules of other institutions.
Scopus ID: 85127565706
2021 IEEE 13th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM), (2021), pp. 1-6
Conference Paper | Published: January 1, 2021
Abstract
Knowledge in Braille is ultimately necessary to maintain learning for the visually impaired. In the Philippines, class attendance has been showing low rates for visually impaired students caused by the shortages of teachers and the absence of the specialized tools intended for teaching them. A proposed solution in addressing this problem is the usage of computers for the automation in the process of the extraction of information in Braille which can facilitate teaching. In recent years, a considerable amount of effort and attention have been devoted to the development of this kind of technology however in languages other than Filipino Braille. Codes in Filipino Braille with its contractions, and even the Filipino language itself has unique features as compared with other languages. In this paper, a system is proposed which uses machine vision in recognizing Filipino Braille codes including one-cell and two-cell contractions. Synthetic Braille images undergo cascade object detection, image processing, extraction of HOG features to develop the three-stage multiclass SVM classifier. Experimental evaluation results reveal a good performance of Filipino Braille classification and translation to texts.