ComTech: Computer, Mathematics and Engineering Applications

Heat Treatment and Its Effect on Tensile Strength of Fused Deposition Modeling 3D-Printed Titanium-Polylactic Acid (PLA)

Mahros Darsin — 2024-11-07

Titanium is a biocompatible metal commonly applied in biomedical fields such as bone and dental implants. Recently, the produced titanium-Polylactic Acid (PLA) filament for 3D printing Fused Deposition Modeling (FDM) technique is easier to operate and affordable. This filament contains less than 20% PLA, which is also biocompatible but hydrophobic and capable of producing inflammation of the surrounding artificial living tissue. Therefore, a heat treatment is needed to reduce or even eliminate PLA. The research aimed to optimize the mechanical properties and biocompatibility of titanium-PLA filaments through heat treatment, demonstrating significant advancements in 3D printing applications for biocompatible materials. A Thermogravimetric Analysis (TGA) was carried out to find out the right temperature for reducing PLA levels. Specimens were heat treated with four temperatures at 100^oC, 160^oC, 190^oC, and 543^oC, and two holding times of 60 and 120 minutes. The mass of the specimens was weighed before and after heat treatment to determine the mass reduction and tested for tensile, micrograph, and fractography observation. The result is a meagre mass reduction. The highest tensile strength of the heat-treated specimen with a heat treatment temperature of 160^oC and a holding time of 60 minutes is 18.310 MPa. However, it is still below the strength of the non-heat treated specimen, 19.890 MPa. Specimens with low tensile strength have a microstructure that shows an uneven distribution of titanium particles. Last, fractography shows porosity in the specimens with the lowest tensile strength.

Model Prediction Using Artificial Neural Network (ANN) to Strengthen Diagnostic Analysis of Diabetes Melitus

Deddy Kurniawan — 2024-11-07

The incidence of Diabetes Mellitus (DM) cases is one of the urgent and increasing health issues every year. Hence, this condition requires high urgency to be handled. The research aimed to develop a prediction model for DM that could be used in general for the purpose of diagnostic analysis of DM cases against suspected individuals. The dataset was sourced from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), which had closely related parameters in diagnostic analysis efforts without favoring certain groups. The targeted contribution was the result of a new prediction model that was specifically tested on the dataset using the Artificial Neural Network (ANN) algorithm. This model was developed through a baseline model that was tested and improved in performance through hyperparameter cross-validation therapy and L1 regularization. The formation of the model architecture through experiments to adjust the conditions of hidden layers and neurons in several configurations results in a model architecture with 8 input parameters. It contains 3 hidden layers with a total of 14, 20, and 26 neurons, with the ReLU activation function on each hidden layer and the Sigmoid activation function on the output part. The second test is carried out on a hyperparameter configuration. It produces maximum performance with a k-fold value of 10 and L1 regularization of 0.0001. The model performance results obtain an accuracy value of 0.947, precision of 0.895, recall of 0.914, and model loss of 0.215

Hierarchical Cluster Analysis Based on Waste Sources in Indonesia in 2022

Syarif Hidayatullah — 2024-11-12

Waste, as a result of human activities, is a complex issue that requires appropriate solutions. With the increasing volume of waste, waste management in Indonesia has become a major challenge. The research examined the waste problem in Indonesia, focusing on analyzing and grouping 311 regencies/cities based on waste sources in 2022. The research also aimed to provide an in-depth understanding of waste characteristics in each region as a basis for designing more effective waste management policies at the regional level. The research applied hierarchical clustering, combining Ward’s method with Euclidean distance analysis. The analysis shows 14 significant clusters with different waste composition characteristics. Interpretation of the cluster results identifies areas with low to high levels of waste. Clusters 1 to 4 have relatively little waste composition, while clusters 5 to 14 have increasing waste levels, with cluster 14 being an area with very high waste levels. The research results are expected to serve as a basis for the government to formulate more targeted and adaptive policies for handling waste in the future. The implications include improving waste management systems, recycling programs, and community education. By understanding the waste composition of each region, the government can implement solutions that suit its needs. The research provides an overview of the waste problem at the regional level in Indonesia and can be the basis for developing more effective policies. In future research, it is recommended to use more accurate and complete waste data in each regency/city for more in-depth results.

Association Analysis Using Apriori Algorithm of GANs-Expanded Student Performance Dataset

Rannie M. Sumacot — 2024-11-12

Traditional datasets are often limited, which can affect the accuracy of analyses. Additionally, the use of students’ real data raises privacy concerns. Generative Adversarial Networks (GANs) offer a solution by generating synthetic data that closely mirrors real-world data without compromising sensitive information. The research explored the application of GANs to enhance student performance datasets by addressing challenges related to data scarcity and privacy in educational research. In the research, GANs were utilized to generate synthetic student performance data. The accuracy of the data was assessed using Mean Absolute Percentage Error (MAPE), with values ranging from 0.004% to 19.92% across various statistical measures and means. These results demonstrated the reliability of the synthetic data, making it suitable for further analysis. The synthetic datasets were then analyzed using the Apriori Algorithm, a well-known method in data mining for discovering significant patterns and relationships. A lower bound minimum support of 0.1 (10%) and a minimum confidence threshold of 0.6 (60%) were applied, ensuring the identification of meaningful associations. The analysis reveals important patterns and relationships among student attributes and behaviors. The research highlights the potential of GANs to advance data-driven educational research. By generating high-quality synthetic data, GANs allow researchers to conduct comprehensive analyses while addressing privacy concerns. The research contributes to the methodological approach to data augmentation in education, offering new opportunities for ethical and robust research.

A Novel Machine Learning for Ethanol and Methanol Classification with Capacitive Soil Moisture (CSM) Sensors

Devina Intan Sari — 2024-11-19

Although Gas Chromatography (GC) is highly accurate, it is costly, highlighting the need for a more affordable method for alcohol detection. Ethanol and methanol have different evaporation rates and dielectric constants, suggesting the potential for classification as an alternative initial step to GC based on differences in dielectric due to evaporation using Capacitive Soil Moisture (CSM) sensors, although it has not been previously attempted. The research aimed to present a novel machine learning for ethanol and methanol classification with CSM sensors. The method involved placing evaporated samples on CSM plates and measuring the change in evaporative dielectric properties over time. The data were then processed using Python, preprocessing data, splitting data, and training various classifiers with key differentiators based on standard deviation, mean, difference, and cumulative summary. Then, model accuracy was evaluated. The research results show that the approach can distinguish between pure ethanol and methanol based on the dielectric differences in each substance's evaporation rate using machine learning training methods with classifiers such as Random Forest, Extra Trees, Gaussian Naive Bayes, AdaBoost, and Logistic Regression with seven folds in cross-validation, L2 regularization, and Newton-Cholesky solver, with accuracies of 96.67%, 96.67%, 96.67%, 93.33%, and 93.33%, respectively. Although the research is limited to the classification of two types of alcohol, the novel approach can classify methanol and ethanol, leading to a potential initial step in determining alcohol content in the future. It can be an alternative to GC with a simpler and more affordable setup using CSM sensors.