Scopus ID: 85096936405
2020 11th IEEE Control and System Graduate Research Colloquium (ICSGRC), (2020), pp. 328-333
Conference Paper | Published: August 1, 2020
Abstract
One way to determine whether or not the student is conscientious in the classroom is by facial expressions. Facial expressions are facial changes in response to a person's internal mental states, thoughts, or social contact. The application of machine learning and computer vision methods have made very useful in area of automated assessment. In this paper, an experimental setup was installed for data collection. The researchers aim to present a new approach of predicting student behavior (attentive or not attentive) based from face recognition during class session. This demonstrate a real-time detection of student behavior. Using deep learning approach, the acquired data utilized the YOLO (you only look once) v3 algorithm in predicting student behavior inside the classroom. The evaluation was created right after the live feed review. Generated models were tested using mAP to decide which model is appropriate for object detection. The mAP (mean average accuracy) is a common measure used to determine the precision of the artifacts being observed. This measure was focused on the following class: high = Attentive and low = Not Attentive. The experimental testing shows that model accuracy is 88.606%. Tests indicate that this method offers reasonable pace of identification and positive outcomes for the measurement of student interest dependent on observable student actions in classroom instruction.
Scopus ID: 85096893634
2020 11th IEEE Control and System Graduate Research Colloquium (ICSGRC), (2020), pp. 323-327
Conference Paper | Published: August 1, 2020
Abstract
Liquefied Petroleum Gas (LPG) is readily available to consumers around the world because LPG is a clean, portable, and efficient source of energy. LPG is currently used by hundreds of millions of people in the world and there are more than 1000 applications. LPG cylinder may leak as a liquid or a gas. If the source of ignition and gas meets LPG can explode or burn. LPG can make the lives of the people in the house endangered and can cause cold burns to the skin. To ensure the safety and security of the consumers using LPG the study aims to develop a device that can detect gas, smoke, and flame from the LPG cylinder and can notify the owner via text messages applying a rule-based approach. The system can display also a warning message and can alarm the owner using a buzzer. The researchers conducted functionality testing with twenty (20) trials in checking the accuracy of the system were obtained a 100% accuracy. This proves that the system is reliable and efficient when using at home by the consumers.
Scopus ID: 85091333737
2020 International Conference on Communication and Signal Processing (ICCSP), (2020), pp. 1054-1058
Abraham T. Magpantay
& Proceso L. FernandezConference Paper | Published: July 1, 2020
Abstract
Image classification in remote sensing is defined by categorizing image pixels or raw data sensed by satellites into a distinct set of labels. In this paper, an improved technique for classifying pixels from satellite images is proposed. The technique makes use of the median value of the pixels in the rectangular neighborhood centered at the given pixel to be classified. A scoring system was developed that compares this median value in relation to the expected median values for each of the different classes. The proposed method was tested on Landsat-8 Operational Land Imager (OLI) bands 1 to 7 images and three index images-Normalized Difference Vegetation Index (NDVI), Normalized Difference Built-up Index (NDBI), and Normalized Difference Water Index (NDWI). The experimental results showed an overall accuracy of 94%, a remarkable improvement from the 84% accuracy of the previous work that uses a distance-based classifier. The obtained results indicate that the proposed method can be a better alternative way to classify images in remote sensing.
Scopus ID: 85089752037
Proceedings of the 2020 3rd International Conference on Robot Systems and Applications, (2020), pp. 82-87
Conference Paper | Published: June 14, 2020
Abstract
The study aims to integrate neural network algorithm that predicts students' vulnerability of not having graduation on time to an adaptive learning management system. Neural network is one of the popular machine learning techniques because of its learning algorithm. The learning algorithm is focused on updating weights of the edges in order to produce minimal mean squared error between actual and predicted values. The integration of this platform could lead to much efficient learning management system as LMS is mainly driven to provide individualized and personalized learning tailored to specific requirements and learning preferences. The neural network algorithm is designed to classify students with learning difficulty so that administrators can formulate remediation and academic support policies.
Scopus ID: 85080932573
2019 International Conference on Computational Intelligence and Knowledge Economy (ICCIKE), (2019), pp. 718-723
Conference Paper | Published: December 1, 2019
Abstract
Medicine intake, as prescribed by physicians and health care providers, is important not only for minimizing the risk of relapse but also to treating conditions and improving one’s overall well-being. However, adherence to a medication routine seems to be a problem for some people which is usually affected by a variety of factors such as hectic day-to-day activity schedules, poor prescription instruction, concurrent intake of multiple medications, and forgetfulness. Medication adherence has been then considered as one of the major medical problems globally. In such cases, a medical device that could alert and remind patients in taking their medicines on time will come in handy. Consequently, this study aimed to design and develop a pocket-sized electronic pillbox device using TFT LCD display, Arduino microcontroller, Piezo Buzzer (for sound notification), Eccentric Rotating Mass (for vibration notification), Lithium Ion battery as power source, and plastic organizer as the main body. The said pillbox device will act as a countermeasure for medication non-adherence particularly by patients under the case of polypharmacy. Thus, this study focused on the design and development of the prototype, hardware testing and system qualification only. Furthermore, this paper is part of a future study where the assessment and measure of device behavior and adherence will be conducted to compare whether the utilization of pillbox device has an impact to the people who are using it.
Scopus ID: 85080942322
2019 International Conference on Computational Intelligence and Knowledge Economy (ICCIKE), (2019), pp. 608-613
Marie Luvett I. Goh
& Joselito Eduard E. GohConference Paper | Published: December 1, 2019
Abstract
In the Philippines, Light Rail Transit (LRT) 1 is one of the most used public mass rapid transports by Filipino commuters in going to their respective destinations in Metro Manila. However, conditions of the trains have been deteriorating over the past years resulting to insufficient numbers of trains to meet the commuter demands during peak hours causing irate passengers, delays in train arrival and uncomfortably crowded stations and trains. Currently, LRT1 implements Passenger Limit Per Platform (PLPP) to regulate load capacity at the station platforms, prevent overloading of trains and congestion at the paid area. But the said scheme is being done manually which is tedious to staff and is prone to error. Thus, this study presents the integration of embedded system and different software applications to manage the crowd of all LRT1 stations platforms and trains intelligently. A Simulation software was developed to populate data to different stations that are relevant during operations in the absence of the station prototype. Integration and acceptance tests showed that all components of the system are functioning accurately according to the predetermined design specifications. The developed system proves to be functionally acceptable in terms of suitability and accuracy, and highly functional in terms of security. Thus, the overall system is functionally acceptable as perceived by the respondents as manifested by the mean rating of 3.28.
Scopus ID: 85080880148
2019 International Conference on Computational Intelligence and Knowledge Economy (ICCIKE), (2019), pp. 581-586
Conference Paper | Published: December 1, 2019
Abstract
This study is concerned with the development of an information system for disaster risk reduction and management in the Philippines. It covered relational and multi-dimensional database designs as well as software applications for disaster preparedness and response combined with Decision Support System. The application highlights the following administrative modules namely community registration with fingerprint biometrics and camera integration, emergency evacuation, search and rescue operation, cash and in-kind donation, evacuation center and disaster event profiling, weather forecast, and private messaging. Moreover, the decision support system highlights the live data consolidation of disaster affected areas and individuals through data visualization and geographic information system. It presents historical information of previous disasters in a multi-dimensional viewpoint from national level to barangay or district level. Finally, the system can dynamically generate predicted list of potential evacuees via Logistic Regression Analysis. The system's response time test revealed a highly acceptable result with latency ranging from 31ms to 419ms. The software quality evaluation in terms of functionality, usability, efficiency, and maintainability proved to be excellent and is highly commendable by the ICT department and operations group of the National Disaster Risk Reduction and Management Council, Office of Civil Defense.
Scopus ID: 85079283298
Proceedings of the 3rd International Conference on Telecommunications and Communication Engineering, (2019), pp. 51-56
Conference Paper | Published: November 9, 2019
Abstract
In the Philippines, according to Philippine Authority of Statistics, there is an imbalance between the student enrollment and student graduation. Almost half of the first-time freshmen full time students who began seeking a bachelor's degree do not graduate on time. The study aims to utilize how Naïve Bayes algorithm - a data classification algorithm that is based on probabilistic analysis - can be used in educational data mining specifically in student graduation. The study is focused on the application of the Naïve Bayes algorithm in predicting student graduation by generating a model that could early predict and identify students who are prone of not having graduation on time, so proper remediation and retention policies can be formulated and implemented by institutions.
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%.
Scopus ID: 85084759812
2019 IEEE 11th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management ( HNICEM ), (2019), pp. 1-6
Conference Paper | Published: November 1, 2019
Abstract
Pulsars enabled astronomers to study neutron stars and verify general relativity under intense gravitational field conditions. However, finding pulsars is not as easy as it seems because most of them have weak pulses that get drowned in the background noise and hence do not get detected. This paper presents a novel way of classifying radio emission patterns collected from a radio telescope whether it is from a pulsar or not through machine learning and genetic algorithm. The dataset was acquired from the High Time Resolution Universe (HTRU) survey two which contains eight numerical features and one target variable describing the pulse profile. Synthetic Minority Oversampling Technique (SMOTE) was applied to the dataset to fix the imbalance between classes. A genetic algorithm library was used to automatically select the best feature preprocessing method, feature selection/reduction technique, machine learning model inside the scikit-learn library, and hyperparameter settings. The genetic algorithm suggested using a single stack and multiple stack classifiers for different sets of features. The optimum level of hyperparameters was also given with the help of the same algorithm. The selected pipelines consistently reported a score of more than 99% in all the evaluation metrics used.