Garuda - Garba Rujukan Digital

IDX30 Stocks Clustering with K-Means Algorithm based on Expected Return and Value at Risk Ahmad Fawaid Ridwan; Subiyanto Subiyanto; Sudradjat Supian
International Journal of Quantitative Research and Modeling Vol 2, No 4 (2021)
Publisher : Research Collaboration Community (RCC)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46336/ijqrm.v2i4.157

Stocks are one of the investment instruments available in the capital market. Several indices show the characteristics of stocks listed on the Indonesia Stock Exchange. IDX30 is one of several indications that show the combined stocks are stocks with large market capitalization, high liquidity, and good fundamentals. The selection of assets to be allocated in the portfolio is an important factor in investing where the purpose of investing is to maximize returns and minimize risk. This study aims to classify stocks that have certain characteristics based on the expected return and value at risk of the stocks incorporated in IDX30 with a clustering algorithm. The clustering algorithm used is the K-Means algorithm. K-Means is a non-hierarchical clustering algorithm by groups each object based on its proximity to the cluster center. The method used in this research is a clustering simulation study using the K-Means algorithm on IDX30 stock data. By identifying the characteristics of the stock based on the characteristics of the cluster formed, it is hoped that it can be considered in choosing the assets to be used in the formation of an optimal portfolio.

SIMULASI PERGERAKAN HARGA SAHAM MENGGUNAKAN MODEL GERAK BROWN GEOMETRIK DENGAN R STUDIO Ahmad Fawaid Ridwan; Rizki Apriva Hidayana; Budi Nurani Ruchjana
Pattimura Proceeding 2021: Prosiding KNM XX
Publisher : Pattimura University

Saham merupakan surat berharga sebagai bukti penyertaan atau kepemilikan individu maupun instansi dalam suatu perusahaaan. Keuntungan berinvestasi saham dapat dilihat dari besarnya return saham. Model matematis dapat diterapkan untuk memodelkan pergerakan harga saham agar investor memiliki pengetahuan untuk memprediksi harga saham di masa mendatang. Salah satu pemodelan yang dapat digunakan untuk melihat pergerakan harga saham yaitu dengan model gerak Brown geometrik. Penelitian ini bertujuan untuk mengetahui tingkat akurasi prediksi pergerakan harga saham menggunakan gerak Brown geometrik. Gerak Brown merupakan sebuah proses stokastik yang bersifat kontinu dan sering disebut sebagai proses Wiener. Gerak Brown dapat dibentuk dari sebuah Random Walk yang simetris, yaitu dengan mencari nilai limit dari distribusi Random Walk tersebut. Model Gerak Brown Geometrik merupakan modifikasi dari gerak Brown dimana perubahan relatifnya berbentuk kombinasi dari pertumbuhan deterministik ditambah dengan perubahan acak yang berdistribusi normal. Metode penelitian menggunakan studi literatur dan studi eksperimental melalui simulasi pada data sekunder berupa data harian harga saham suatu perusahaan dari tanggal 4 Mei 2020 sampai 30 April 2021. Simulasi data menggunakan gerak Brown geometrik dengan R Studio menunjukkan bahwa data return saham berdistribusi normal serta menghasilkan prediksi pergerakan harga saham dengan tingkat akurasi yang baik. Hal ini ditunjukkan dengan nilai rata-rata MAPE dari 20 kali percobaan simulasi menggunakan R Studio, yaitu sebesar 13,904 %. Oleh karena itu, model gerak Brown geometrik dapat digunakan oleh para investor atau manager investasi untuk memprediksi pergerakan harga saham suatu perusahaan dalam rentang waktu tertentu

IDX30 Stocks Clustering with K-Means Algorithm based on Expected Return and Value at Risk Ahmad Fawaid Ridwan; Subiyanto Subiyanto; Sudradjat Supian
International Journal of Quantitative Research and Modeling Vol. 2 No. 4 (2021): International Journal of Quantitative Research and Modeling
Publisher : Research Collaboration Community (RCC)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46336/ijqrm.v2i4.157

Stocks are one of the investment instruments available in the capital market. Several indices show the characteristics of stocks listed on the Indonesia Stock Exchange. IDX30 is one of several indications that show the combined stocks are stocks with large market capitalization, high liquidity, and good fundamentals. The selection of assets to be allocated in the portfolio is an important factor in investing where the purpose of investing is to maximize returns and minimize risk. This study aims to classify stocks that have certain characteristics based on the expected return and value at risk of the stocks incorporated in IDX30 with a clustering algorithm. The clustering algorithm used is the K-Means algorithm. K-Means is a non-hierarchical clustering algorithm by groups each object based on its proximity to the cluster center. The method used in this research is a clustering simulation study using the K-Means algorithm on IDX30 stock data. By identifying the characteristics of the stock based on the characteristics of the cluster formed, it is hoped that it can be considered in choosing the assets to be used in the formation of an optimal portfolio.

Spectral Energy-Based Adaptive Homomorphic Filtering for Astronomical Image Enhancement Ahmad Fawaid Ridwan; Iqbal Robiyana
Jurnal Matematika, Statistika dan Komputasi Vol. 23 No. 1 (2026): September 2026
Publisher : Department of Mathematics, Hasanuddin University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20956/bfpvzv53

Astronomical images often suffer from low contrast, uneven illumination, dark backgrounds, and sensor-induced noise, which can obscure faint celestial structures and reduce the effectiveness of subsequent image analysis. Conventional homomorphic filtering can improve illumination and contrast by attenuating low-frequency components and enhancing high-frequency details, but its use of a fixed cutoff frequency limits its adaptability to images with different spectral characteristics. This study proposes a Spectral Energy-Based Adaptive Homomorphic Filtering (SEAHF) method that estimates the cutoff frequency from the spectral energy ratio between low- and high-frequency components of each input image. The proposed mechanism modifies only the cutoff frequency while retaining the conventional Gaussian homomorphic filtering framework and fixed remaining filter parameters. The method was evaluated on 60 astronomical images from the SDSS, DECaLS, and Pan-STARRS surveys, selected from common sky coordinates available across the three sources, and compared with Contrast Limited Adaptive Histogram Equalization (CLAHE), Conventional Homomorphic Filtering (CHF), and Low-Light Image Enhancement via Illumination Map Estimation (LIME). Performance was evaluated using Structural Similarity Index Measure (SSIM), entropy, Lightness Order Error (LOE), and processing time. SEAHF consistently achieved the lowest LOE across the three datasets, with an overall LOE of 532570.5, while obtaining competitive SSIM (0.5714), entropy (6.0646), and processing time (0.009374 s). Although CLAHE and CHF achieved higher SSIM and CLAHE required less processing time, SEAHF provided the strongest illumination-order preservation while maintaining computational cost comparable to CHF. The results provide initial empirical evidence that spectral-energy-based adaptive cutoff estimation can improve illumination-order preservation while maintaining competitive structural similarity and computational cost.

Transformasi Intensitas Adaptif dengan Preservasi Urutan Tingkat Kecerahan pada Citra Cahaya Rendah Ahmad Fawaid Ridwan; Muhammad Herlambang Prakasa Yudha
JIMAT: Jurnal Ilmiah Matematika Vol 7 No 3 (2026): Juli-September 2026
Publisher : Program Studi Pendidikan Matematika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.63976/jimat.v7i3.1557

Citra dengan kondisi pencahayaan rendah umumnya memiliki kontras yang rendah, detail objek yang kurang terlihat, serta rentan terhadap derau sehingga dapat menurunkan kualitas visual dan performa analisis lanjutan. Metode peningkatan kualitas citra berbasis transformasi intensitas dapat meningkatkan visibilitas area gelap, tetapi pemilihan parameter transformasi yang tidak adaptif dapat menyebabkan peningkatan intensitas yang berlebihan (over-enhancement) dan perubahan hubungan tingkat kecerahan (light order). Penelitian ini mengusulkan Adaptive Light-Order Preserving Intensity Transformation (ALPIT) sebagai metode peningkatan kualitas citra berbasis transformasi intensitas adaptif. Metode yang diusulkan membangun fungsi transformasi piecewise menggunakan adaptive anchor point yang ditentukan berdasarkan rata-rata intensitas, variansi, dan entropi ternormalisasi sehingga proses transformasi dapat disesuaikan dengan karakteristik distribusi intensitas citra. Kinerja ALPIT dievaluasi secara visual dan kuantitatif pada tiga citra dengan kondisi pencahayaan rendah dan dibandingkan dengan Gamma Correction (GC) serta Contrast Limited Adaptive Histogram Equalization (CLAHE) menggunakan metrik SSIM dan LOE. Pada tiga citra yang diuji, ALPIT memperoleh nilai rata-rata SSIM sebesar 0,6717, lebih tinggi dibandingkan GC (0,3992) dan CLAHE (0,4755), serta nilai rata-rata LOE sebesar 104,5199, lebih rendah dibandingkan GC (127,7193) dan CLAHE (209,2536). Hasil visual menunjukkan peningkatan visibilitas pada area gelap dengan perubahan intensitas pada area terang yang relatif terkendali. Hasil ini menunjukkan bahwa ALPIT merupakan pendekatan awal yang berpotensi meningkatan citra low-light, dengan hasil evaluasi yang pada penelitian ini masih terbatas pada tiga citra grayscale dan dua metode pembanding klasik.

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