Claim Missing Document
Check
Articles

Found 4 Documents
Search

Performance Comparison of Ant Colony Optimization and Artificial Bee Colony in Solving the Capacitated Vehicle Routing Problem Setyawan, Deva Putra; Lianingsih, Nestia; Saputra, Moch Panji Agung
International Journal of Global Operations Research Vol. 5 No. 4 (2024): International Journal of Global Operations Research (IJGOR), November 2024
Publisher : iora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47194/ijgor.v5i4.339

Abstract

The Capacitated Vehicle Routing Problem (CVRP) is a combinatorial optimization problem widely applied in logistics and supply chain management. It involves determining the optimal routes for a fleet of vehicles with limited capacity to serve a set of customers with specific demands while minimizing travel costs. This study compares the performance of two popular metaheuristic algorithms, Ant Colony Optimization (ACO) and Artificial Bee Colony (ABC), in solving the CVRP. The research implements both algorithms on standard benchmark datasets, evaluating solution accuracy and computational efficiency. Simulation results indicate that ACO tends to excel in finding high-quality solutions, particularly for problems with high complexity, whereas ABC demonstrates superior computational efficiency on small- to medium-scale datasets. A detailed analysis of algorithm parameters was also conducted to understand their impact on the performance of both methods. This study provides valuable insights into the strengths and limitations of each algorithm in the context of CVRP and paves the way for the development of hybrid approaches in the future.
Indirect Methods for Personalized Mean-CVaR Portfolio Optimization Setyawan, Deva Putra; Salih, Yasir
Operations Research: International Conference Series Vol. 6 No. 2 (2025): Operations Research International Conference Series (ORICS), June 2025
Publisher : Indonesian Operations Research Association (IORA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47194/orics.v6i2.385

Abstract

This study presents a reformulation of the Personalized Mean-CVaR model into an unconstrained optimization problem, which is then solved using iterative methods, including steepest descent and Newton’s method. The reformulation introduces challenges related to feasibility region checking, convexity of the feasible set and objective functions, and the use of Lagrange multipliers to handle constraints. Additionally, Taylor expansion is utilized to approximate the objective function in each iteration. The research evaluates the effectiveness of iterative optimization techniques in solving the Personalized Mean-CVaR problem, while addressing key challenges in convergence and stability of the solution.
Indonesian Banking Stock Portfolio Optimization Based on Ridge Regression Prediction Saputra, Moch Panji Agung; Setyawan, Deva Putra; Wahid, Alim Jaizul
International Journal of Business, Economics, and Social Development Vol. 6 No. 2 (2025)
Publisher : Rescollacom (Research Collaborations Community)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46336/ijbesd.v6i2.1064

Abstract

The Indonesian stock market in the banking sector is a popular investment instrument with high return potential but faces market volatility and global economic uncertainty that requires adaptive and data-driven portfolio management strategies. Traditional asset allocation strategies such as equal weighting or based on historical performance have limitations in dynamic market conditions, while the application of machine learning, especially Ridge Regression, in stock return prediction and portfolio optimization in the Indonesian market has not been widely explored. This study aims to build an integrated pipeline for portfolio prediction and optimization using Ridge Regression on Indonesian banking stocks. Methods: Daily closing price data of five major banking stocks (BBRI, BBCA, BMRI, BBNI, BBTN) for the period 2015-2025 are used with technical indicators of moving average and rolling standard deviation as input features. The Ridge Regression model is trained using TimeSeriesSplit cross-validation to predict daily returns, then the prediction results are integrated into the Mean-Variance optimization framework to maximize the Sharpe ratio. The Ridge Regression model shows excellent predictive performance with an average R² of 0.9986, MAE of 0.000466, and RMSE of 0.000720. The Ridge-based portfolio strategy achieves identical performance to the historical optimal strategy with an annualized return of 10.64% and a Sharpe ratio of 0.4705, significantly outperforming the equal-weight strategy (return of 6.63%, Sharpe ratio of 0.2562). A practical implementation simulation with IDR 100 million funds shows feasible execution with less than 1% deviation from the optimal weights. Ridge Regression is proven to be effective in capturing the return pattern of Indonesian banking stocks and enables superior portfolio performance when integrated with modern portfolio theory, providing investors with a robust and data-driven approach to portfolio optimization in emerging markets.
Implementation of Machine Learning Model for Pest Classification in Rice Plants Saputra, Moch Panji Agung; Setyawan, Deva Putra; Dwiputra, Muhammad Bintang Eighista
International Journal of Research in Community Services Vol. 6 No. 3 (2025): International Journal of Research in Community Service (IJRCS)
Publisher : Research Collaboration Community (Rescollacom)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46336/ijrcs.v6i3.957

Abstract

Rice cultivation is a cornerstone of food security in agrarian countries like Indonesia, yet it remains highly vulnerable to pest infestations that can severely impact crop productivity. Manual identification of pests is time-consuming and error-prone, especially when pest species exhibit similar morphological characteristics. This study aims to implement and evaluate the performance of four classical machine learning algorithms Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Random Forest (RF), and Logistic Regression (LR) for classifying rice pests based on image data. The dataset, derived from Kaggle’s “Rice Pest Detection Dataset,” includes 12 pest classes and underwent a series of preprocessing steps: grayscale conversion, image resizing to 128×128 pixels, feature extraction using Histogram of Oriented Gradients (HOG), label encoding, and class balancing via SMOTE. The experimental setup used 80% of the data for training and 20% for testing. Performance was evaluated using precision, recall, F1-score, and confusion matrices. Among the four models, SVM achieved the most consistent and robust performance, with F1-scores reaching up to 0.98 in several pest classes and an overall balanced classification across the dataset. Random Forest followed closely, particularly excelling in distinguishing classes such as Rice Water Weevil and Yellow Rice Borer, achieving F1-scores of 0.99 and 0.96 respectively. In contrast, KNN showed signs of overfitting, with extreme precision-recall imbalances, while LR was more stable but less accurate in separating visually similar classes like Rice Stem Fly and Thrips. Visual analysis of correct and incorrect predictions revealed that classes 7 (Rice Stem Fly) and 11 (Thrips) were consistently misclassified across all models, likely due to high visual similarity.