ejournal.nusamandiri.ac.id (STMIK Nusa Mandiri)
Not a member yet
1504 research outputs found
Sort by
COMPARISON OF MACHINE LEARNING ALGORITHMS FOR SENTIMENT ANALYSIS OF DIGITAL IDENTITY APPLICATION USERS
Get PDFIn the rapidly evolving digital era, the Population Identity Application (IKD) plays a crucial role in streamlining civil administration processes in Indonesia, allowing easier and faster access to population services. This study aims to explore the application of machine learning algorithms in analyzing user responses to the IKD application. Three popular algorithms: Support Vector Machine (SVM), K-Nearest Neighbors (K-NN), and Naïve Bayes were chosen to classify sentiment from 1301 user reviews on the Google Play Store into positive and negative categories. After performing data preprocessing such as tokenization and stemming, hyperparameter optimization was conducted using GridSearchCV to enhance classification accuracy. The research results indicate that the SVM algorithm, optimized with hyperparameters, including the use of the rbf kernel and a parameter value of C = 1, achieved the highest accuracy of 85.60%, making it the most effective method for sentiment classification of the IKD application. These findings provide valuable insights for the government and developers in refining the features and performance of IKD, contributing to the efficiency and security of digital administration in Indonesia. Furthermore, this study opens opportunities for further development that is more responsive to user needs and expectations in the future
IMPROVING STUNTING CLASSIFICATION PERFORMANCE USING COMBINATION SMOTE TECHNIQUE AND ARTIFICIAL NEURAL NETWORK ALGORITHM
Get PDFChild development is at the core of the nation's future. However, there are still serious problems that hinder optimal child growth, one of which is stunting. Stunting is a condition that has become a global concern in the context of public health and development. The use of deep learning algorithms has great potential to overcome the problem of stunting classification. The ratio of stunting handling is still a problem due to imbalance data. Classification algorithms such as ANN will experience a decrease in performance when faced with unbalanced classes, this makes it difficult to take action for early diagnosis of stunting. Synthetic Minority Oversampling Technique (SMOTE) is used to balance the failure data in diagnosis. The results showed that the combination of the SMOTE oversampling technique can improve the ability of the ANN algorithm model to accurately classify stunted or minority populations. The accuracy, precision, recall, and F1-Score values of this study are 0.90, 0.85, and 0.95, respectively. The results of (MCC) obtained a value of 0.73, and (G-Mean) of 0.86 before applying SMOTE and the results after applying SMOTE MCC of 0.84 and G-Mean of 0.92. This indicates that the minority group, namely stunted toddlers, can be predicted well. The implementation of the combination of SMOTE and ANN has proven successful in classifying imbalance stunting data, so this research can be used as a reference for future research to handle unbalanced data
ELECTRIC BICYCLE SELECTION SYSTEM USING MULTI CRITERIA DECISION MAKING
Get PDFIndonesia government has targeted that by 2025 electric vehicle production has reached 20%. Many electric bicycle products have been sold and used in Indonesia, both domestically made and imported. The most important technical considerations in choosing an electric bicycle are battery type and motor used. These two components affect distance traveled by electric bicycles even further. Many brands offered by manufacturers with various advantages and disadvantages will make consumers are confused about choosing a quality and durable electric bicycle. From these problems, one solution is to create a mobile-based decision support system (DSS) that can assist in making decisions according to predetermined criteria. This research aims to produce recommendations for electric bicycle brands that can assist potential consumers in choosing electric bicycles. The combination of the Analytical Hierarchy Process (AHP) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) methods is used in order to produce more objective recommendations. In this research uses seven criteria that will be processed in AHP, namely price, maximum load, battery capacity, motor power, charging time, mileage and speed. There are nine brands that are used as alternatives in the TOPSIS calculation. From AHP and TOPSIS calculation, following recommendations for electric bicycles is obtained: the first order is the Ofero with a preference value of 1, next order are with the same value 0.5806, Indobike and Exotic, and the last recommendation is Jarvis with a value of 0.095
COMPARISON OF SVM AND NAÏVE BAYES CLASSIFIER ALGORITHMS ON STUDENT INTEREST IN JOINING MSIB
Get PDFMachine learning (ML) is a branch of artificial intelligence (AI) that deals with the development of systems capable of learning from data to make predictions or decisions without being explicitly programmed. In this study, we conducted an analysis of students' interest in the Internship and Certified Independent Study Program (MSIB) in the context of the Independent Campus Learning policy. The method used is a survey by distributing questionnaires to students of Amikom Purwokerto University in the MSIB batch 5 in year 2023. The results of this study can provide understanding and predictions about students' interest in the MSIB program based on relevant variables, such as study program, semester, cumulative grade point average (GPA), semester credit system (SKS), and previous work experience. The research results indicate that GPA and Study Program greatly influence students' interest in MSIB. The Naïve Bayes algorithm yielded an accuracy of 0.6875 on the training data and 0.25 on the testing data, with a confusion matrix of (0, 1, 0; 0, 1, 2; 0, 0, 0). Meanwhile, the Support Vector Machine (SVM) algorithm yielded an accuracy of 0.4375 on the training data and 0.75 on the testing data, with a confusion matrix of (0, 1; 0, 3). The machine learning model developed in this study is expected to help predict students interest based on new data provided, thus supporting decision-making in optimizing the MSIB program
EFFECTIVITY IMPROVEMENT OF HYBRID PROJECT MANAGEMENT WATER-SCRUM-FALL WITH SIX SIGMA IMPLEMENTATION
Get PDFIn software development projects there is continuous development aimed at increasing the efficiency and effectiveness of the team in providing software quality and customer satisfaction. The problem is that many projects are planned using a waterfall approach by clients, therefore some tools are needed to balance this situation. This research aims to evaluate the effectiveness of hybrid methodologies in software development by uncovering the use of a combination of Waterfall and Agile Scrum methodologies with the application of Six Sigma. This hybrid methodology was deemed suitable for combining the plan- and contract-based characteristics of Waterfall with the flexibility and rapid iteration of Agile Scrum. The use of Six Sigma is used to focus on change, assist in systematically identifying and correcting process problems, and process improvement. In research methods, sample teams run different methodologies on similar software projects. Hybrid project management is carried out by applying the Waterfall approach in planning and contracts. At the same time, each phase in Waterfall is iterated using Agile Scrum to ensure flexibility and adaptability. The research results found that this hybrid method can increase team efficiency, reduce development cycle time, detect higher defects in each sprint, increase the final quality of the software, and finally increase the Sigma Index with a team comparison of σ=3.22 and σ= 3.11 and higher compared to teams that only use Agile-Scrum. In conclusion, the integration of Waterfall, Agile Scrum, and Six Sigma can be an effective strategy to face the challenges of modern software development
LEARNING AUTOMATA-BASED AODV ROUTING PROTOCOL TO IMPROVE V2V COMMUNICATION IN URBAN TRAFFIC SIMULATION
Get PDFNetwork congestion, packet loss, and high latency in the AODV routing protocol are significant obstacles to achieving reliable vehicle-to-vehicle (V2V) communication. Consequently, an update to the AODV protocol is necessary. This research proposes the Learning Automata-based AODV (LA-AODV) routing protocol to address these issues. The LA-AODV protocol incorporates learning automata into the routing protocol by considering speed, acceleration, and x and y coordinates. The communication quality index with the nearest vehicles is measured before selecting a set of relay nodes until the maximum estimated time is reached. The primary objective of this study is to enhance the performance of V2V communications by reducing network congestion, packet loss, and latency. The results demonstrate that LA-AODV achieves a maximum packet delivery ratio (PDR) improvement of 4.0% and a throughput of up to 56.50 kbps, surpassing the performance of both AODV and DSDV protocols. These findings indicate the potential of LA-AODV to optimize V2V communications, thereby significantly improving transportation safety and efficiency. The research contributes to the field by providing a novel solution to enhance V2V communication quality in urban traffic scenarios, offering significant benefits in reduced latency, increased reliability, and overall better network performance
K-BEST SELECTION UNTUK MENINGKATKAN KINERJA ARTIFICIAL NEURAL NETWORK DALAM MEMPREDIKSI RANGE HARGA PONSEL
Get PDFDetermining the price of a mobile phone that will be released to the market cannot be based on assumptions alone. This problem can be overcome by utilizing machine learning. In this study, what is predicted is not the exact price, but rather the price range of a cellphone based on the specifications that are its attributes. In machine learning, the Deep Learning ANN model will be used to predict the price range of a mobile phone. To understand the relationship between features and labels, the Univariate feature selection method SelectKBest is used which will calculate the correlation value between features and labels. In this study, the best performance was obtained from the ANN model with feature selection and hyperparameter tuning, the evaluation of performance metrics obtained the highest accuracy of 97.5%. Experiments were conducted by building several models to compare until there was one model that performed well in processing training and validation data. Model evaluation is presented using confusion metrics with various types of performance metrics: accuracy, precision, recall and f1-score. This study also aims to evaluate the effectiveness of the SelectKBest feature selection method in improving model accuracy and testing various hyperparameter configurations to obtain the best performance
IMPLEMENTASI METODE WATERFALL DAN SYSTEM USABILITY SCALE TESTING PADA APLIKASI FISIOTERAPI PASIEN BPJS
Get PDFThe background of the problem in this study is that Physiotherapy Services at Permata Hati Hospital for BPJS patients are carried out scheduling or physiotherapy protocol data by writing on the form paper provided and then given to the patient. Besides being given to the patient, the form is also stored by the physiotherapy poly section. The problem that often occurs is that the physiotherapy poly officer must also rewrite the BPJS patient's physiotherapy protocol data in the ledger as data for the next physiotherapy schedule. If you want to find physiotherapy protocol data, it is difficult to do because you have to look one by one in the ledger. The purpose of this research is to make it easier for poly officers to process physiotherapy protocol data in a computerized manner through the design of BPJS patient physiotherapy protocol data applications at Permata Hati Hospital by applying the waterfall method and System Usability Scale (SUS). The results showed that in blackbox testing, the results were obtained in accordance with what was expected starting from the login form test to the patient data input form. Furthermore, testing using the SUS method obtained an average value of 70.75 with a total of 10 respondents and a total of 10 statements. There are 3 components of the System Usability Scale method, namely Acceptable, Grade Scale and Adjective, each component of the results is still at a good value and the level of application is quite comfortable to be used by users
REBRANDING UMKM CATERING MELALUI DESAIN PACKAGING DAN PROMOSI MEDIA SOSIAL
Get PDFThe problem faced in the Mbok Ndut Catering and Bu Sri Catering catering businesses is that there is no label on the packaging and does not have media as a promotion, so it requires rebranding the packaging design in the form of labels and promotion of the business on social media. The purpose of this community service activity is to provide solutions to the problems being faced. The method used is to provide training in the form of tutorials on how to manage Instagram accounts and how to operate the Google Maps application. The results of this community service are to help the two MSME players create a logo design, create an official Instagram account, create a content design for Instagram posts, create a packaging design in the form of labels, add the location of the Catering business address through the Google Maps Application, train or teach the two MSME players, manage Instagram by posting the menu that we provide with interesting content, and how to operate the Google Maps Application. Thus, in the future it can make it easier for the two MSME players to introduce their business to the wider community
PENDAMPINGAN DIGITALISASI SEKOLAH SMPIT AJIMUTU GLOBAL INSANI TAMBUN UTARA-BEKASI MENUJU MANAJEMEN SEKOLAH BERBASIS DIGITAL
Get PDFIn the digital era, the application of information technology in schools is crucial to enhance the efficiency and quality of educational management. This Community Service (PkM) activity aims to assist Sekolah Menengah Pertama Islam Terpadu (SMPIT) Ajimutu Global Insani Tambun Utara-Bekasi in implementing digital-based school management. The main objective of this activity is to optimize the school's management system through a tailored application that meets the school's needs, called DigiMS (Digital School Management). DigiMS will be built web-based which will be hosted and use a domain name, and a mobile-based application will be created which will be submitted to the Playstore. The method used includes stages of analyzing the existing school management system, developing the DigiMS application, providing training for staff and teachers on how to use the application, and monitoring and evaluating the application's implementation. Data was collected through direct observations at the partner school, interviews with staff, and efficiency measurements before and after the implementation of the digital system. The results of this activity showed a significant increase in operational efficiency, especially in student data management, financial administration, and internal and external communication. The main conclusion from this program is that adopting digital-based management through DigiMS can improve the effectiveness of school management and ease staff workloads in their daily tasks. Additionally, this program has a positive impact on the school's readiness to face the challenges of the digital era. This program is part of the 2024 Fiscal Year PKM Grant from Kemdikbudristek-DRTPM, conducted by lecturers and students from Universitas Nusa Mandiri