Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi)
Not a member yet
1071 research outputs found
Sort by
Logistic Regression Using Hyperparameter Optimization on COVID-19 Patients’ Vital Status
Get PDFThis study aims to classify COVID-19 patients based on the results of their hematology tests. Hematology test results have been shown to be useful in identifying the severity and risk of COVID-19 patients. Specifically, this study focuses on classifying COVID-19 patients based on their vital status, namely Deceased and Alive. The dataset used in this study contains four variables: white blood cells (WBC), neutrophils (NEU), lymphocytes (LYM), and Neutrophil Lymphocyte Ratio (NLR). Logistic Regression algorithm was used to solve the problem, and hyperparameter optimization was implemented to obtain the best model performance. The objective of this study was to build the best parameter in classifying the patients’ vital status. The proposed model achieved an accuracy score of 78%, which is the best performance among the tested models. The results of this study provide a key component for decision making in hospitals, as it provides a way to quickly and accurately identify the vital status of COVID-19 patients. This study has important implications for managing the COVID-19 pandemic and should be of interest to researchers and practitioners in the field.
This study aims to classify COVID-19 patients based on the results of their hematology tests. Hematology test results have been shown to be useful in identifying the severity and risk of COVID-19 patients. Specifically, this study focuses on classifying COVID-19 patients based on their vital status, namely Deceased and Alive. The dataset used in this study contains four variables: white blood cells (WBC), neutrophils (NEU), lymphocytes (LYM), and Neutrophil Lymphocyte Ratio (NLR). Logistic Regression algorithm was used to solve the problem, and hyperparameter optimization was implemented to obtain the best model performance. The objective of this study was to build the best parameter in classifying the patients’ vital status. The proposed model achieved an accuracy score of 78%, which is the best performance among the tested models. The results of this study provide a key component for decision making in hospitals, as it provides a way to quickly and accurately identify the vital status of COVID-19 patients. This study has important implications for managing the COVID-19 pandemic and should be of interest to researchers and practitioners in the field
Improving the Accuracy of C4.5 Algorithm with Chi-Square Method on Pure Tea Classification Using Electronic Nose
No full textTea is one of the plantation products within the Ministry of Agriculture of the Republic of Indonesia, which plays an essential role as a mainstay commodity that boosts the Indonesian economy. Each type of tea has different properties, and the aroma of each type of tea can measure the quality of the tea. The human sense of smell is still very limited in classifying pure types of tea. Therefore, a device is needed to help measure the aroma of tea from an electronic nose. The devices attached to several gas sensors help humans take data from the smell of pure tea and calculate the value of each type of tea to test datasets with data mining algorithms. This study uses the C4.5 algorithm as a classification method with advantages over noise data, missing values, and handling variables with discrete and continuous types. Meanwhile, Chi-square is used to perform attribute severing in the data preprocessing process to increase the accuracy of dataset testing. Testing a pure tea dataset with four whole attributes, namely CO2, CO, H2, and CH4, using the C4.5 algorithm resulted in an accuracy of 93.65% and an increase in the accuracy performance of the C4.5 algorithm by 94.27% with dataset testing using Chi-Square feature selection with the two highest value attributes.Tea is one of the plantation products within the Ministry of Agriculture of the Republic of Indonesia, which plays an essential role as a mainstay commodity that boosts the Indonesian economy. Each type of tea has different properties, and the aroma of each type of tea can measure the quality of the tea. The human sense of smell is still very limited in classifying pure types of tea. Therefore, a device is needed to help measure the aroma of tea from an electronic nose. The devices attached to several gas sensors help humans take data from the smell of pure tea and calculate the value of each type of tea to test datasets with data mining algorithms. This study uses the C4.5 algorithm as a classification method with advantages over noise data, missing values, and handling variables with discrete and continuous types. Meanwhile, Chi-square is used to perform attribute severing in the data preprocessing process to increase the accuracy of dataset testing. Testing a pure tea dataset with four whole attributes, namely CO2, CO, H2, and CH4, using the C4.5 algorithm resulted in an accuracy of 93.65% and an increase in the accuracy performance of the C4.5 algorithm by 94.27% with dataset testing using Chi-Square feature selection with the two highest value attributes
The Examination of the User Engagement Scale (UES) in Small Medium Enterprise Social Media Usage: A Survey-Based Quantitative Study
Get PDFSocial networks have proven to be an essential marketing tool for the success of any product, service, or business. User participation affects the increase in revenue gain and creates long-term profit. The User Engagement Scale (UES) is one of the tools developed to measure user engagement and has been used in various digital domains. The UES intends to compute six dimensions of engagement: aesthetic appeal, perceived usability, focused attention, novelty, felt involvement, and endurability. This study investigates and verifies the three-factor structure of the UES. We used PCA to perform the analysis. The original data will be reanalyzed using UES, which consists of 220 valid responses. The result shows that the UES examination indicates good reliability in three factors. Factor 1 encompasses the feeling of involvement (FI), aesthetic appeal (AE), novelty (NO), and endurability (EN). Factor 2 aggregates the perceived usability (PU) elements. Factor 3 pertains to focused attention (FA) items. Our findings indicate that the User Engagement Scale is a valuable and suitable measurement tool for assessing user engagement in the context of social media within small and medium enterprises
Identifikasi Penyakit Tanaman Pisang Melalui Citra Daun Pisang Menggunakan Metode CNN Dengan Model ResNet50 dan VGG-19
No full textIdentify banana plant diseases using machine learning with the CNN method to make it easier to identify diseases in banana plants through leaf images. It employs the CNN method, incorporating ResNet50 because ResNet50 is one of the best models and a suitable model for the data set used, and the VGG-19 model is used because VGG-19 was one of the winning models of the 2014 ImageNet Challenge and is a model that also fits the data set used. The research objectives encompass data set processing, model architecture development, evaluation, and result reporting, all aimed at improving disease identification in banana plants. The ResNet50 model achieved impressive 94% accuracy, with 88% precision, 91% recall, and an F1 score of 89%, while the VGG-19 model demonstrated strong performance with 91% accuracy, surpassing previous research and highlighting the effectiveness of these models in identifying banana plant diseases through leaf images. In conclusion, the exceptional accuracy positions it as the preferred model for CNN-based disease identification in banana plants, offering significant advances and insights for agricultural practices. Future research opportunities include exploring alternative CNN models, architectural variations, and more extensive training datasets to improve disease identification accuracy.Pisang ialah salah satu tanaman yang tumbuh di daerah tropis yaitu dari Kawasan Asia Tenggara. Selain itu pisang adalah tanaman yang dapat berbuah sepanjang tahun atau tidak berbuah musiman, sehingga memiliki nilai ekonomi yang lumayan tinggi, dan kekuatan untuk dapat terbuka luasnya pasar penjualan serta permintaan yang terus meingkat. Nilai ekspor pisang di Indonesia adalah 30.372.955 kg dengan nilai ekspor US$14.609.697 pada tahun 2020 dan terdapat 3 provinsi penghasil pisang terbesar di Indonesia yaitu, Provinsi Jawa Timur sebesar 2.116.974 ton, provinsi Jawa Barat sebanyak 1.220.174 ton dan provinsi Lampung sebanyak 1.209.545 ton. Penyakit yang menyerang pada daun yang dinamakan bercak daun atau biasa dikenal penyakit Sigatoka, gejala pertama dari penyakit sigatoka ini biasanya bintik-bintik berwarna kuning muda dalam jangka waktu beberapa hari bintik-bintik akan menambah 1 hingga 2 cm panjang dan menjadi warna cokelat di bagian daun, jika dibiarkan akan mengakibatkan kegagalan panen yang dapat merugikan para petani. Metode CNN dengan model ResNet50 yang diusulkan untuk skema pengujian 1 dan model VGG-19 untuk skema pengujian model 2, dengan dataset yang diambil melalui situs Kaggle yang berjudul Banana Leaf Dataset. Dataset yang diambil memiliki 4 kelas klasifikasi yaitu, healthy, cordana, sigatoka, dan pestalotiopsiz lalu ukuran data akan di ubat menjadi 224 x 224 dan melakukan split data dengan rasio 80% data pelatihan dan 20% untuk data validasi. Hasil pengujian dari ke 2 skema mendapatkan hasil 94% accuracy dari skema pengujian model ResNet50 dimana lebih besar dari pada skema pengujian model VGG-19 sebesar 91%
Abstractive and Extractive Approaches for Summarizing Multi-document Travel Reviews
Get PDFTravel reviews offer insights into users' experiences at places they have visited, including hotels, restaurants, and tourist attractions. Reviews are a type of multidocument, where one place has several reviews from different users. Automatic summarization can help users get the main information in multi-document. Automatic summarization consists of abstractive and extractive approaches. The abstractive approach has the advantage of producing coherent and concise sentences, while the extractive approach has the advantage of producing an informative summary. However, there are weaknesses in the abstractive approach, which results in inaccurate and less information. On the other hand, the extractive approach produces longer sentences compared to the abstractive approach. Based on the characteristics of both approaches, we combine abstractive and extractive methods to produce a more concise and informative summary than can be achieved using either approach alone. To assess the effectiveness of abstractive and extractive, we use ROUGE based on lexical overlaps and BERTScore based on contextual embeddings which it be compared with a partial approach (abstractive only or extractive only). The experimental results demonstrate that the combination of abstractive and extractive approaches, namely BERT-EXT, leads to improved performance. The ROUGE-1 (unigram), ROUGE-2 (bigram), ROUGE-L (longest subsequence), and BERTScore values are 29.48%, 5.76%, 33.59%, and 54.38%, respectively. Combining abstractive and extractive approach yields higher performance than the partial approach
Forecasting the Magnitude Category Based on The Flores Sea Earthquake
Get PDFEarthquakes are a phenomenon that is still a mystery in terms of predicting events, one of which is the magnitude. As technology develops, there are many algorithms that can be used as approaches in earthquake forecasting. In the context of magnitude forecasting, the application of GaussianNB, Random Forest, and SVM has the potential to reveal these patterns and relationships in the data. With the six main phases of this research, namely data acquisition, data pre-processing, feature selection, model training, forecast result evaluation, and performance analysis, this study is expected to contribute to the development of more accurate and effective earthquake forecasting methods. From these results we first obtain the result that the GaussianNB model has a relatively simple and fast method in training its model. However, the weakness lies in the assumption of a Gaussian distribution, which may not always suit the complex and diverse characteristics of earthquake data. Second, Random Forest, this method can increase accuracy and overcome the overfitting problem that occurs when forecasting magnitudes. In contrast to GaussianNB, it tends to result in models with greater complexity and requires more time to compute. The third option is SVM, which has both benefits and drawbacks that must be taken into account. The capacity of SVM to separate data that has both linear and nonlinear separation is one of its key advantages; nevertheless, the main drawback is that it is sensitive to hyperparameter adjustments
Iris Recognition Using Hybrid Self-Organizing Map Classifier and Daugman’s Algorithm
Get PDFOne of the neural network algorithms that can be used in iris recognition is self-organizing map (SOM). This algorithm has a weakness in determining the initial weight of the network, which is generally carried out randomly, which can result in a decrease in accuracy when an incorrect determination is made. The solution that is often used is to apply a hybrid process in determining the initial weight of the SOM network. This study takes an approach using the cosine similarity equation to determine the initial weight of the network SOM in order to increase recognition accuracy. In addition, the localization process needs to be carried out to limit the area of the iris image being studied so that it is easy for the recognition process to be carried out. The method proposed in this study for iris recognition, namely hybrid SOM and Daugman’s algorithm, has been tested on several people by capturing the iris of the eye using a digital camera. The captured eyes have been localized first using the Daugman’s algorithm, and then the image features were extracted using the GLCM and LBP methods. In the final stage of the study, an iris recognition comparison test was performed, and the results obtained an accuracy of 85.50% using the proposed method and an accuracy of 73.50% without performing a hybrid process on the SOM network.
One of the neural network algorithms that can be used in iris recognition is self-organizing map (SOM). This algorithm has a weakness in determining the initial weight of the network, which is generally carried out randomly, which can result in a decrease in accuracy when an incorrect determination is made. The solution that is often used is to apply a hybrid process in determining the initial weight of the SOM network. This study takes an approach using the cosine similarity equation to determine the initial weight of the network SOM in order to increase recognition accuracy. In addition, the localization process needs to be carried out to limit the area of the iris image being studied so that it is easy for the recognition process to be carried out. The method proposed in this study for iris recognition, namely hybrid SOM and Daugman’s algorithm, has been tested on several people by capturing the iris of the eye using a digital camera. The captured eyes have been localized first using the Daugman’s algorithm, and then the image features were extracted using the GLCM and LBP methods. In the final stage of the study, an iris recognition comparison test was performed, and the results obtained an accuracy of 85.50% using the proposed method and an accuracy of 73.50% without performing a hybrid process on the SOM network
IoT Microcontroller Application Prototype as Data Transceiver from Network to USB Device
Get PDFTelecommunications technology in the early 2000s until now has experienced a rapid increase. Starting with complex devices to microcomputer devices that are able to connect to the network. With the development of networking technology, it will leave behind non-network-support devices that can still be used. The existence of previous research on "The interface between the IoT microcontroller (ESP32) and the Max3421e USB Host" can be taken advantage of developing a device that can facilitate a non-support-network device into a support-network electronic device. So that non-support-network electronic devices do not become electronic waste and can also become a device that can be used in the present. In this study, a prototype of a data receiver from WiFi was designed, then the data that has been received is reorganized into rows of data that are ready to be sent to non-support-network electronic devices. This developed tool uses an ESP32 IoT microcontroller connected to the USB Host max3421e which has been packaged in the form of a USB host shield module using SPI protocol communication. The result obtained is that data from the network can be sent correctly to the USB Host max3421e via the ESP32 microcontroller.Telecommunications technology in the early 2000s until now has experienced a rapid increase. Starting with complex devices to microcomputer devices that are able to connect to the network. With the development of networking technology, it will leave behind non-network-support devices that can still be used. The existence of previous research on "The interface between the IoT microcontroller (ESP32) and the Max3421e USB Host" can be taken advantage of developing a device that can facilitate a non-support-network device into a support-network electronic device. So that non-support-network electronic devices do not become electronic waste and can also become a device that can be used in the present. In this study, a prototype of a data receiver from WiFi was designed, then the data that has been received is reorganized into rows of data that are ready to be sent to non-support-network electronic devices. This developed tool uses an ESP32 IoT microcontroller connected to the USB Host max3421e which has been packaged in the form of a USB host shield module using SPI protocol communication. The result obtained is that data from the network can be sent correctly to the USB Host max3421e via the ESP32 microcontroller
Indonesian Hate Speech Detection Using Bidirectional Long Short-Term Memory (Bi-LSTM)
Get PDFAbstract
Social media is a platform that allows users to express themselves freely including spreading hate speech content. The government has issued the regulation in the UU ITE to handle and prevent hate speech on social media. The research was also conducted using the Bi-LSTM to classify the text into hate speech or not. Another research was purposed to detect hate speech and its categories using Bi-GRU. However, the performance of the model Bi-GRU is still lower than Bi-LSTM with an accuracy of 86.44% and 96.44%. Therefore, this study aims to build a model that can detect hate speech and its categories. The research offers Bi-LSTM as a classification model and IndoBERT as a tokenization model. The dataset used is a public dataset containing 13 thousand tweets. As a result, the best model obtained is using 20 epochs, 192 batch sizes, 1 layer Bi-LSTM with 40 nodes, and applying class weighing in the optimization process. The pre-train model from IndoBERT that is used to support the performance of the model in classifying is "indobenchmark/indobert-large-p2". The performance given by the purposed model is very good with an average accuracy, precision, and recall of 97.66%, 96.50%, and 85.25%.Abstract
Social media is a platform that allows users to express themselves freely including spreading hate speech content. The government has issued the regulation in the UU ITE to handle and prevent hate speech on social media. The research was also conducted using the Bi-LSTM to classify the text into hate speech or not. Another research was purposed to detect hate speech and its categories using Bi-GRU. However, the performance of the model Bi-GRU is still lower than Bi-LSTM with an accuracy of 86.44% and 96.44%. Therefore, this study aims to build a model that can detect hate speech and its categories. The research offers Bi-LSTM as a classification model and IndoBERT as a tokenization model. The dataset used is a public dataset containing 13 thousand tweets. As a result, the best model obtained is using 20 epochs, 192 batch sizes, 1 layer Bi-LSTM with 40 nodes, and applying class weighing in the optimization process. The pre-train model from IndoBERT that is used to support the performance of the model in classifying is "indobenchmark/indobert-large-p2". The performance given by the purposed model is very good with an average accuracy, precision, and recall of 97.66%, 96.50%, and 85.25%
Penerapan E-Test Pada Mata Pelajaran Bahasa Inggris Menggunakan Model View Controller
No full textThe problem that occurs in one of the high schools is that there is no online-based English exam system, and they still use paper for their assessments. This makes teachers unable to give exams to students if they do not attend school. The purpose of this research is to create an online exam system to facilitate exam administration using the Model View Controller (MVC) to assess online English exams. The analytical method uses the System Usability Scale (SUS), to collect data using an electronic test (e-Test).Unifield Modeling Language (UML), a visual modeling technique used in designing and building object-oriented software for E-Test applications.The results showed that the use of the E-test application in a comparison class in English subjects could improve student learning outcomes to a higher level, from a score of 81.7 to 90.9. Compared to the control class, there was a slight increase, namely an increase in value from 71.5 to 75.8. This proves that the E-test application is used effectively by students for English subjects.Penelitian ini bertujuan untuk mengkaji penggunaan Model View Controller (MVC) untuk mengevaluasi ujian online. Sumber data adalah mata pelajaran Bahasa Inggris untuk siswa SMA Adabiah Padang. Pengumpulan data dilakukan melalui uji elektronik (e-Test) dengan menggunakan metode analisis System Usability Scale (SUS). Sistem ujian online menggunakan Unifield Modeling Language (UML), salah satu metode pemodelan visual yang digunakan dalam perancangan dan pembuatan perangkat lunak berorientasi objek. Hasil penelitian sangat baik untuk kategori kata sifat dan baik untuk kategori dapat diterima. Berdasarkan hasil tersebut dapat disimpulkan bahwa model ini sangat cocok untuk dikembangkan dalam proses penilaian pembelajaran. Kajian sistem ujian online ini bertujuan untuk mempermudah pelaksanaan ujian. Hasil penelitian ini akan memberikan manfaat bagi setiap guru dan siswa untuk memaksimalkan penggunaan teknologi komputer. Sehingga nantinya akan membantu para siswa dan guru untuk melaksanakan ujian di sekolah masing-masing