Garuda - Garba Rujukan Digital

Neural Network for Predicting Dining Experiences at Restaurant X Anggara Putra, I Wayan Kintara; Santiyuda, Kadek Gemilang
Jurnal Sistem Informasi dan Komputer Terapan Indonesia (JSIKTI) Vol 7 No 4 (2025): June
Publisher : INFOTEKS (Information Technology, Computer and Sciences)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33173/jsikti.217

This study explores the application of neural networks to predict dining experiences at Restaurant X, utilizing a combination of customer feedback, operational data, and sales transactions. The goal is to enhance restaurant management through accurate predictions of customer satisfaction and operational performance. Customer reviews, sentiment analysis, and operational data were processed using natural language processing (NLP) and time-series analysis to prepare the data for neural network training. The model's performance was evaluated using metrics such as accuracy, precision, recall, and F1-score, and it was compared with traditional machine learning techniques like logistic regression and decision trees. The results demonstrate that neural networks outperform traditional algorithms in predicting customer sentiment and dining experiences. This study highlights the potential of deep learning to provide valuable insights into customer behavior, enabling restaurants to improve service personalization, marketing strategies, and operational efficiency. Future research can focus on expanding the dataset and exploring more advanced deep learning models to further enhance prediction accuracy and applicability in the hospitality industry.

Developing Trading Strategies for Doge Coin with Reinforcement Learning Anggara Putra, I Wayan Kintara
Jurnal Sistem Informasi dan Komputer Terapan Indonesia (JSIKTI) Vol 6 No 3 (2024): March
Publisher : INFOTEKS (Information Technology, Computer and Sciences)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33173/jsikti.232

Cryptocurrency trading, particularly with highly volatile assets like Dogecoin, presents significant challenges due to rapid price fluctuations and external factors such as social media sentiment and speculative trading behaviors. This study proposes reinforcement learning (RL)-based trading strategies to address these complexities. RL, an advanced machine learning approach, enables dynamic adaptation to market conditions by optimizing sequential decisions for maximum cumulative rewards. Using historical market data and technical indicators, RL agents were trained and evaluated in simulated trading environments. Performance metrics, including profitability, risk-adjusted returns, and robustness under varying market conditions, demonstrate that RL-based strategies outperform traditional methods by capturing non-linear dependencies and responding effectively to delayed rewards. The results highlight the ability of RL to adapt to market volatility and optimize trading outcomes. However, the study acknowledges limitations, including the exclusion of external sentiment data and restricted testing across diverse market scenarios. Future research should integrate external data sources, such as sentiment and macroeconomic indicators, conduct real-time market testing, and explore applications to multi-asset portfolios to improve generalizability and robustness. This research contributes to the intersection of machine learning and financial markets, showcasing RL’s potential to address cryptocurrency trading challenges and offering pathways for more adaptive and robust trading strategies.

A Hybrid Approach to Chili Price Classification Using Ensemble Methods Anggara Putra, I Wayan Kintara
Jurnal Sistem Informasi dan Komputer Terapan Indonesia (JSIKTI) Vol 7 No 1 (2024): September
Publisher : INFOTEKS (Information Technology, Computer and Sciences)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33173/jsikti.241

This study proposes a hybrid machine learning approach for predicting chili prices, integrating ensemble methods such as Random Forest, Gradient Boosting, and XGBoost to enhance forecasting accuracy. By analyzing historical price data, the model identifies key features, including day and value, as significant predictors. The hybrid model demonstrates superior performance in capturing non-linear patterns and seasonal variations compared to individual machine learning techniques. Evaluation metrics such as Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) validate the model’s effectiveness in handling market volatility. The findings highlight the potential of advanced machine learning techniques in agricultural price forecasting, offering reliable and actionable insights for farmers, traders, and policymakers. This approach not only addresses challenges in market prediction but also provides a scalable framework for future enhancements, such as incorporating additional variables like weather and supply chain factors. By bridging the gap between data-driven analysis and practical application, this research contributes to stabilizing agricultural markets and supporting informed decision-making processes.

Classifying UKT Fee Relief Eligibility Using K Nearest Neighbors Algorithm Anggara Putra, I Wayan Kintara; Agastyar Priatdana, Gde Yoga
Jurnal Sistem Informasi dan Komputer Terapan Indonesia (JSIKTI) Vol 5 No 3 (2023): March
Publisher : INFOTEKS (Information Technology, Computer and Sciences)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33173/jsikti.251

This research develops a K-Nearest Neighbors (KNN)-based classification model to determine the eligibility of students for Tuition Assistance (UKT) based on socio-economic factors, including parental income, family size, parental occupation, number of dependents, and housing conditions. The goal is to automate the process of identifying students eligible for financial aid, enhancing both the efficiency and fairness in resource allocation. The model was trained using a dataset consisting of both categorical and numerical features, with the target variable being binary: "Eligible" (1) or "Not Eligible" (0) for UKT relief. The KNN model achieved an overall accuracy of 92%, with strong performance in predicting the "Eligible" class. However, the "Not Eligible" class showed lower performance, particularly in terms of recall and F1-score, suggesting the presence of class imbalance. To address this issue, techniques such as class balancing, resampling, or adjusting KNN parameters are suggested to improve the model's ability to correctly classify minority instances. Additionally, exploring ensemble methods like Random Forest or XGBoost may provide more robust results. This study highlights the importance of addressing class imbalance and using appropriate evaluation metrics beyond accuracy when building classification models for imbalanced datasets.

Implementation of the TOPSIS Method for a Decision Support System in Recommending Tourist Destinations in Tabanan Anggara Putra, I Wayan Kintara; Agastyar Priatdana, Gde Yoga
Jurnal Sistem Informasi dan Komputer Terapan Indonesia (JSIKTI) Vol 6 No 1 (2023): September
Publisher : INFOTEKS (Information Technology, Computer and Sciences)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33173/jsikti.262

Tourism development plays an important role in stimulating regional economic growth, particularly in areas with diverse natural and cultural attractions such as Tabanan Regency in Bali. However, visitors often experience difficulties in selecting destinations that match their preferences due to the presence of multiple decision factors and scattered informational resources, making destination decisions less systematic and potentially inconsistent. This situation highlights the need for a methodical decision support mechanism capable of evaluating tourist destinations based on multiple criteria. Motivated by this issue, this study implements the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method as part of a Decision Support System designed to recommend suitable tourist destinations in Tabanan. The system evaluates nine destinations based on eight criteria, which include accessibility, attractiveness, facility availability, cleanliness, cost, popularity, safety, and visitor density, and applies weight values determined through expert judgment. The evaluation results show that Jatiluwih Rice Terrace has the highest ranking with a closeness coefficient of 0.679126, followed by Ulun Danu Beratan and Tanah Lot, indicating that heritage value and environmental management strongly contribute to recommendation outcomes. The model provides transparent ranking reasoning and can support tourists, planners, and local tourism administrators in making informed decisions. Future development may involve expanding the destination dataset, integrating real-time visitor data, and deploying the system as a mobile application to improve personalization and accessibility.

Title

Found 5 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 5 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 5 Documents
Search