Journal of Applied Statistics and Data Science

Front Matter Vol 1 No 1 (2024): Journal of Applied Statistics and Data Science

Suci Astutik — Thu, 28 Mar 2024 00:00:00 +0000

Quadratic Nonlinear Path Analysis Modeling on Simulation Data

Erlinda Citra Lucki Efendi — Thu, 28 Mar 2024 00:00:00 +0000

This study aims to analyze quadratic nonlinear paths using the WLS method, perform hypothesis testing, and determine the best model between various sample sizes in the simulated data. The data used in this study is simulated data, where the data is applied to three exogenous variables, one endogenous variable, and one mediating variable with a correlation of 0.9 at various sample sizes (n = 100, 300, and 500). The results of this study indicate that there is variation in the various sample sizes that can affect the smoothness of the curve. The larger the sample size value, the better the resulting curve. The results of testing the linear parameter hypothesis show that there is a significant relationship between X1 and Y1, X2 and Y1, X3 and Y1, and X1 and Y2 at a correlation of 0.9 with a sample size of 100, 300, and 500. The best model is obtained in quadratic nonlinear path analysis with the largest R² value (95%) compared to the coefficient of determination of other models. The originality of this study is by using a quadratic nonlinear path analysis model and the use of simulation data which is rarely studied by other researchers.

Support Vector Machine for Sentiment Analysis of PT. Paragon Technology and Innovation (Case Study of Brand Make Over and Emina Product Users on Female Daily Page – Beauty Review)

Vety Bhakti Lestari; Dini Amalia — Thu, 28 Mar 2024 00:00:00 +0000

Support Vector Machine (SVM) is one of the classification models in Supervised Learning that is commonly used to classify user sentiment towards certain products or services. Facial care products are widely used by all circles, especially Gen-Z. The aim of this study is to obtain sentiment from the specified product reviews and apply the SVM method to predict sentiment classification. The products analyzed consisted of two Emina products, namely sunscreen and face wash and three Make Over products, namely eyebrow pencil, blush, and lipstick. The sentiment results of the five products showed that Make Over blush on products received the most positive sentiment at 98%, while Emina Sunscreen products received the least, with only at 66%. The SVM model in this study showed a good performance in making predictions with accuracy on all five products above 80%. In addition, the level of accuracy of the SVM model in classifying that the data is included in the positive class on the five products is also good because it has a recall value of > 75%. The results of this study can be used as an evaluation for PT. Paragon Technology and Innovation to continuously improve product quality and SVM models can be used to predict classification in other studies.

Development Structural Equation Modelling with Second Order Measurement Model and Mixed Scale Data

Retno Ayu Cahyoningtyas — Thu, 28 Mar 2024 00:00:00 +0000

This research will develop structural equation modeling involving latent variables with a second order measurement model, as well as mixed measurement scales using simulation data. Applied to simple structural modeling which has one exogenous variable, one mediating endogenous variable, and one endogenous dependent variable. Based on the results of the analysis, if the coefficient of the measurement model and the structural model is greater, the error variance is 0.1, causing the correlation between indicators in each variable to be stronger. Another effect caused by the greater closeness of the coefficient of the measurement model and the structural model is that the greater the percentage of diversity generated in the measurement model. The results of the inner model analysis show that there is an insignificant relationship, namely the relationship between the Y1 variable and the Y2 variable but for other path relationships it shows significant. The total coefficient of determination for the second order SEM shows that the greater the closeness of the coefficient of the measurement model and the structural model in the error variance condition of 0.1, the greater the R²value.

Agile-Scrum Implementation in The Development of Kampus Merdeka Information System

Nenni Mona Aruan, Albet Matthew Best Nainggolan, Timothy Sipahutar — Thu, 28 Mar 2024 00:00:00 +0000

Merdeka Belajar Kampus Merdeka (MBKM) program is a program initiated by the Minister of Education and Culture to encourage students to master various disciplines for their preparation in the job market. At the Institut Teknologi Del, the MBKM program is also offered to students. However, its implementation is not yet fully automated and utilizes various different platforms. This research aims to design and develop the Independent Internship & Study Implementation Information System in the Merdeka Campus at the Institut Teknologi Del using the Agile-Scrum approach. The Scrum method is an iterative software development approach that is flexible to changes. The implementation of Agile-Scrum in the development of the MBKM information system helps achieve goals and automate existing systems. This research involves Scrum activities such as sprint planning, daily scrum, sprint review, and sprint retrospective to produce an information system that meets the requirements. The Scrum product backlog is used as a reference during system development throughout the sprint. The development of the information system through agile-scrum has received feedback during the sprint review and has been functionally tested using several testing methods.

Development of Seemingly Unrelated Regression Analysis with Dummy Variables: Modeling the Relationship of Student Learning Outcome Indicators

Muhamad Liswansyah Pratama — Thu, 28 Mar 2024 00:00:00 +0000

This research aims to develop a seemingly unrelated regression model with dummy variables to determine the relationship between factors that influence student learning outcomes in each subject which includes Religious Education, Citizenship, Indonesian, English, Mathematics, Sciences, Social Science, Cultural Arts, Physical Education and Crafts. This research is applied sciences using primary data obtained through a list of questions (questionnaire) given to class IX students at SMP Negeri 42 Batam. The predictor variables of this research include students' creativity, motivation, interests, talents, support from parents, friends and school where interests are the intercept dummy for this research. The results of the research show that there is a positive relationship between interest and the value of each subject and parental support in 9 subjects, but there is a negative relationship between school support in 8 subjects. Creativity and motivation have no relationship in every subject. The regression model obtained was good because 99.6% of the response variables could be explained by the predictor variables. It is hoped that this research will be used as evaluation material for students, parents and schools to improve learning outcomes where class IX students will of course pursue higher education the following year.

Comparison of K-Nearest Neighbor and Support Vector Machine for Sentiment Analysis of the Second COVID-19 Booster Vaccination

Salma Fadhila Sari, Imelda Salsabila — Thu, 28 Mar 2024 00:00:00 +0000

K-Nearest Neighbor (K-NN) and Support Vector Machine (SVM) are classification methods commonly used in sentiment analysis. Sentiment analysis can be applied to analyze the opinion of social media specially on platforms like Twitter about a specific topic. The purpose of this research is to determine and compare the classification accuracy of the K-NN and SVM One Against One algorithm in classifying the sentiments of the public regarding the implementation of the second Covid-19 booster vaccination on Twitter. The data used in this research consists of tweet from July 29th 2022 to February 15th 2023 that contain aspects of adverse events following immunization and perception of vaccination effectiveness. The result of this study obtained that there were 1,576 tweets containing positive sentiments, 169 containing neutral sentiments, and 424 containing negative sentiments. The data is divided into training data and testing data with a ratio of 80%:20%. The classification accuracy obtained for K-NN with k=17 is accuracy of 85.48%, precision of 78.64%, and recall of 85.48%, while for SVM, the accuracy is 84.33%, precision of 78.34%, and recall of 84.33%. Based on the classification obtained by K-NN with k=17, it is better at classifying sentiment regarding the second Covid-19 booster vaccination.

Keywords: K-Nearest Neighbor, Sentiment Analysis, Support Vector Machine, The Second Covid-19 Booster Vaccination