Jurnal Sains, Nalar, dan Aplikasi Teknologi Informasi

Deep Learning Based LSTM Model Hyperparameter Testing to Predict the Number of Road Accidents

Joko Siswanto, Benny Daniawan, Haryani Haryani, Pipit Rusmandani — Thu, 25 Jul 2024 00:00:00 +0000

Many have used the prediction of the number of road accidents, but it is still rare to find those who use and test prediction models that are not suitable. Predictive models that have been used to predict road accidents have proven successful, but have not provided model testing with data that is different from the deep learning approach. The LSTM model test is proposed to be tested with 5 different datasets from Kaggle and 3 hidden layer variations. The test results of the LSTM model are that with variations of 4 hidden layers it can achieve higher accuracy results than those without hidden layers and 2 hidden layers. The results are obtained from stability with the lowest average MSLE value and relatively balanced average time. Deep learning-based LSTM model testing was carried out to ensure and prove the stability of the model for predicting the number of road accidents in the future. Stakeholders can predict the number of road accidents using the resulting prediction model.

Redesign User Interface at PT. Budi Jaya Banjarindo Using UCD Method

Doddy Ariansyah; Irving Vitra Paputungan — Sun, 28 Jul 2024 00:00:00 +0000

Budi Jaya Banjarindo operates in the machining and manufacturing sector and has a production system (make to order), which means the product is made according to consumer requests or design. This company uses machine equipment such as lathes, CNC machines, CNC Plasma Cutting, milling machines, hobbing and so on. Evaluation and redesign of the user interface system using Figma tools which will later be used by the workshop to optimize online product marketing. The system was designed using the UCD method. The final design result is a prototype which will then be assessed or tested using the SUS method to obtain an assessment from users. The SUS Score obtained was 65, based on these results it is included in the C range (range 60-70). These results indicate that the UI evaluation results are of good quality. The conclusion of redesigning the interface by adding several additional features, successfully meets user needs and the quality of test results is good.

Optimizing an Expert System for Diagnosing a Depression Disorder Using a Case Based Reasoning Method

Septian Rico Hernawan, Nur Azmi Ainul Bashir; Ifan Hakim — Wed, 03 Jul 2024 00:00:00 +0000

According to data from the World Health Organization (WHO), 3.7% of the population in Indonesia experiences depression. Depression can impact both the mental and physical conditions of an individual. WHO reports that every year, approximately 800,000 people die by suicide, with depression being one of the causes. Depression treatment is handled by a professional, and in the field, the process of diagnosing depression disorders is still generally done manually by them. This creates many opportunities for errors, despite the fact that each level of depression disorder requires different handling. Inadequate treatment can hinder the patient's recovery and may potentially worsen their condition. A precise and efficient method is needed to diagnose depression disorders. An expert system can reduce the risk of errors that occur with manual calculations. The implemented case-based reasoning method can classify depression disorders. Testing was conducted using 30 datasets as initial knowledge, with 20 sample data points for testing, randomly selected from the population through questionnaires. The classification accuracy for depression disorders reached up to 90%.

Clustering Junior Schools in Implementing Smart School Using The K-Means in Pekanbaru

Aida Nisa, M. Khairul Anam, Helda Yenni, Parlindungan Kudadiri, Gunadi — Wed, 03 Jul 2024 00:00:00 +0000

The purpose of this research is to determine the readiness of schools in implementing the Smart School system through various stages. One of the concepts of a Smart City involves integrating information and communication technology into the learning process at every school to create Smart Schools. However, not all schools are ready to implement this technology because it requires suitable technology to support the quality of teaching and learning. Another issue is the absence of information systems that can facilitate administrative tasks and the teaching and learning process. The use of the K-Means method is beneficial for clustering schools based on their stages, characteristics, and readiness to implement the Smart School system. This helps identify schools with the highest level of readiness. This research demonstrates that the use of K-Means can identify school readiness based on the established stages related to the Smart School system. It also can pique students' interest in developing and boosting the school's reputation as the best technology-based school.

Smartphone Device Monitoring System Using Google Family Link (D’Paragon Housekeeping and Cleaning Service Case Study)

Muhammad Mustofa, Iwan Ady Prabowo, Hendro Wijayanto — Wed, 03 Jul 2024 00:00:00 +0000

D'Paragon Kost is an exclusive boarding house based in Yogyakarta and has approximately 38 branches. Spread throughout Indonesia and in each branch there are employees who serve as Housekeeping and Cleaning Services. At each Housekeeping and Cleaning Service, facilities are provided in the form of a Smartphone device which is useful for reporting scheduled activities. Using a smartphone that is not intended can cause problems with the smartphone. After the IT team investigated the branch location, there were applications that were outside the company's operational standards. So precautions are taken so that smartphone facilities cannot be used to access applications outside of standard procedures. There are several types of applications that can control smartphone devices. One of these applications is Google Family Link. The aim is to remotely control the Housekeeping and Cleaning Service smartphone and limit the performance of the smartphone itself. As a data collection sample, the author used 5 (five) days of data from 4 (four) Housekeeping and 1 (one) Cleaning Service with different assignment locations and jobs. The data used is daily application usage data in minutes and standard application data from D'Paragon Kost. From the analysis of application usage behavior in each branch, the number of working hours does not affect the lack of access to certain applications. If this behavior is outside the company's Standard Operating Procedures (SOP), it can reduce employee performance. This is shown by the existence of the Cleaning Service branch which has been using the YouTube application for longer than other branches. So monitoring each branch in the use of smartphones with Google Family is very effective in knowing employee behavior at work. The company can also provide recommendations regarding the use and access of smartphones according to their intended use.

Clustering Analysis of Chess Portable Game Notation Text

Feri Wijayanto — Thu, 04 Jul 2024 00:00:00 +0000

Chess is a game that requires a high level of intelligence and strategy. Generally, in order to understand complex move patterns and strategies, the expertise of chess masters is required. With the rapid development in the field of machine learning, the digitization of chess game recordings in Portable Game Notation (PGN) format, and the availability of large and widely accessible data, it is possible to apply machine learning techniques to analyze chess games. This research studies the use of text clustering algorithms, specifically hierarchical clustering and K-means clustering, to categorize chess games based on their moves. We extracted 100 chess games that use certain openings such as French Defence, Queen's Gambit Declined, and English Opening. In the implementation of hierarchical clustering, single, average, and complete linkage methods are used. As a result, our findings show that hierarchical clustering with single linkage is less effective. On the other hand, the average and complete linkage methods, as well as K-means clustering, successfully identify clusters corresponding to the original openings. Notably, K-means clustering showed the highest accuracy in clustering chess games. This research highlights the potential of machine learning techniques in uncovering strategic patterns in chess games, paving the way for deeper insights into game strategies.