Garuda - Garba Rujukan Digital

Analisis Data Time Series Menggunakan LSTM (Long Short Term Memory) Dan ARIMA (Autocorrelation Integrated Moving Average) Dalam Bahasa Python. Adhitio Satyo Bayangkari Karno
ULTIMA InfoSys Vol 11 No 1 (2020): Ultima InfoSys : Jurnal Ilmu Sistem Informasi
Publisher : Universitas Multimedia Nusantara

Abstract — This study aims to predict time series data by using two methods, the first method commonly used is statistics Autocorrelation Integrated Moving Average (ARIMA model) and the second method which is relatively new, namely machine learning Long Short Term Memory (LSTM). Before the data is processed by both methods, data cleaning and data optimization are carried out. Data optimization is a transformation process to eliminate elements of trends and variations from data. The transformation consists of 7 results of a combination from Log processes, Moving Average (MA), Exponential Weigh Moving Average (EWMA), and Differencing (Diff). The seven processes are each used in the ARIMA and LSTM processes. So that 14 predictions will be obtained (7 from the ARIMA process and 7 from the LSTM process). From the 14 prediction results obtained the smallest RMSE value for ARIMA is 2% and the smallest RMSE value for LSTM is 1%. The results of this study using 7 combinations of transformation processes, can increase the level of accuracy of predictions from ARIMA and LSTM. Where the accuracy of LSTM learning machines by using Telkom's stock data has higher accuracy than ARIMA.

Deep Learning - Prediksi dan Resiko Investasi Enam Saham Bank di Indonesia Nurdina Widanti; Tri Surawan; Adhitio Satyo Bayangkari Karno; Aji Digdoyo; Nani Kurniawati; Rachmat Nursalam; Harini Agusta
Jurnal Teknologi Vol 10, No 1 (2022): Jurnal Teknologi
Publisher : Universitas Jayabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31479/jtek.v10i1.194

Stock prediction and risk level are important for investors, but this ability can only be done by experts and takes a long time. The rapid development of technology today demands that decisions be made quickly and precisely. Deep Learning (DL) is one of the Artificial Intelligence (AI) methods that are able to analyze and predict stock values accurately, in real-time, and without much human intervention. Analysis of risk and correlation between stocks by calculating daily returns using the moving average (MA) method. Dataset of 6 bank shares obtained from yahoo-finance, namely Bank Central Asia (BBCA), Bank Rakyat Indonesia (BBRI), Bank Mandiri (BMRI), Bank Nasional Indonesia (BBNI), Bank Rakyat Indonesia Syariah (BRIS), and Bank Tabungan Negara (BBTN). The volume of share sales increased significantly only in BBRI and BBNI shares, although 5 bank shares (except BRIS) experienced price increases. The highest correlation occurred between BBNI and BMRI shares with a value of 97%, 92% between BMRI and BBCA shares, and 91% between BBNI and BBCA. Analysis of risk and expected return shows that BRIS has the highest risk and expected return of 0.042245 and 0.002986, respectively. BBCA shares have the lowest risk and expected return at 0.015392 and 0.000695, respectively. The results show that future predictions have decreased, namely BBRI, BBNI, and BBTN, and rose for BBCA, BMRI, and BRIS stocks.

Optimizing Random Forest Models for Early Detection of Defects in Steel Tri Surawan; Adhitio Satyo Bayangkari Karno; Widi Hastomo; Reza Fitriansyah; Ahmad Eko Saputro; Indra Bakti
Journal of Informatic and Information Security Vol. 5 No. 1 (2024): Juni 2024
Publisher : Program Studi Informatika, Fakultas Ilmu Komputer Universitas Bhayangkara Jakarta Raya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31599/amqxgn78

In the manufacturing sector, steel plate defects are a severe issue that may result in significant losses for a company's finances and image. The purpose of this study is to evaluate how well three machine learning algorithms detect steel plate flaws. The accuracy, area under the ROC curve (ROC-AUC), and Log-Loss of the method were used to assess its performance using a dataset that was downloaded from www.kaggle.com. Based on the findings, the Random-Forest algorithm performed best overall, having the lowest Log-Loss of 0.9327, an accuracy of 0.6722, and an AUC value of 0.9222. Research using other algorithms is still very open to be carried out to get better results. Research utilizing other algorithms is still very much open to be conducted in order to get better outcomes.

Improved Banking Customer Retention Prediction Based on Advanced Machine Learning Models Linda Wahyu Widianti; Adhitio Satyo Bayangkari Karno; Hastomo, Widi; Aryo Nur Utomo; Dodi Arif; Indra Sari Kusuma Wardhana; Deon Strydom
Indonesian Journal of Information Systems Vol. 7 No. 2 (2025): February 2025
Publisher : Program Studi Sistem Informasi Universitas Atma Jaya Yogyakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24002/ijis.v7i2.10364

The quick growth of the banking sector is reflected in the rise in the number of banks. In addition to the intense competition among banks for new customers, efforts to keep existing ones are essential to minimizing potential losses for the company. To ascertain whether customers will leave the bank or remain customers, this study will employ churn forecasts. A 1,750,036-customer demographic dataset, which includes data on bank customers who have left or are still customers, is used in the training process to compare five machine learning technology models in order to investigate the improvement of binary classification prediction accuracy. These models are Decision Tree, Random Forest, Gradient Boost, Cat Boost, and Light Gradient Boosting Machine (LGBM). According to the study's results, LGBM performs better than the other four models since it has the highest recall and accuracy and the fewest False Negatives. The LGBM model's corresponding accuracy, precision, recall, f1 score, and AUC are 0.8789, 0.8978, 0.8553, 0.8758, and 0.9694. This demonstrates that, in comparison to traditional methods, machine learning optimization can produce notable advantages in churn risk classification. This study offers compelling proof that sophisticated machine learning modeling can revolutionize banking industry client retention management.

A Breakthrough in Viral Pneumonia Detection: Unveiling Insights with ResNet-152 Widi Hastomo; Adhitio Satyo Bayangkari Karno; Nani Kurniawati; Harini Agusta
Journal of Informatic and Information Security Vol. 4 No. 2 (2023): Desember 2023
Publisher : Program Studi Informatika, Fakultas Ilmu Komputer Universitas Bhayangkara Jakarta Raya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31599/1zcjsb83

Viral pneumonia is one of the most serious health issues. The key problem in providing early detection and rapid mitigation through the use of chest X-ray imaging has become the ability to identify accurately. The ResNet-152 convolutional neural network approach will be used in this study to predict viral pneumonia. The input dataset was obtained from Kaggle.com. The accuracy findings from this investigation obtained a substantial value, namely 0.99, indicating that the model used performed admirably. The model used can efficiently distinguish between the viral pneumonia dataset and other datasets. It is intended that the findings of this study will be used to inform early decisions in related medical sectors.

Diagnosa COVID-19 Chest X-Ray Menggunakan Arsitektur Inception Resnet Adhitio Satyo Bayangkari Karno; Dodi Arif; Indra Sari Kusuma Wardhana; Eka Sally Moreta
Journal of Informatic and Information Security Vol. 2 No. 1 (2021): Juni 2021
Publisher : Program Studi Informatika, Fakultas Ilmu Komputer Universitas Bhayangkara Jakarta Raya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31599/abbs9m42

The availability of medical aids in adequate quantities is very much needed to assist the work of the medical staff in dealing with the very large number of Covid patients. Artificial Intelligence (AI) with the Deep Learning (DL) method, especially the Convolution Neural Network (CNN), is able to diagnose Chest X-ray images generated by the Computer Tomography Scanner (C.T. Scan) against certain diseases (Covid). Inception Resnet Version 2 architecture was used in this study to train a dataset of 4000 images, consisting of 4 classifications namely covid, normal, lung opacity and viral pneumonia with 1,000 images each. The results of the study with 50 epoch training obtained very good values for the accuracy of training and validation of 95.5% and 91.8%, respectively. The test with 4000 image dataset obtained 98% accuracy testing, with the precision of each class being Covid (99%), Lung_Opacity (97%), Normal (99%) and Viral pneumonia (99%).

Improved Banking Customer Retention Prediction Based on Advanced Machine Learning Models Linda Wahyu Widianti; Adhitio Satyo Bayangkari Karno; Hastomo, Widi; Aryo Nur Utomo; Dodi Arif; Indra Sari Kusuma Wardhana; Deon Strydom
Indonesian Journal of Information Systems Vol. 7 No. 2 (2025): February 2025
Publisher : Program Studi Sistem Informasi Universitas Atma Jaya Yogyakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24002/ijis.v7i2.10364

The quick growth of the banking sector is reflected in the rise in the number of banks. In addition to the intense competition among banks for new customers, efforts to keep existing ones are essential to minimizing potential losses for the company. To ascertain whether customers will leave the bank or remain customers, this study will employ churn forecasts. A 1,750,036-customer demographic dataset, which includes data on bank customers who have left or are still customers, is used in the training process to compare five machine learning technology models in order to investigate the improvement of binary classification prediction accuracy. These models are Decision Tree, Random Forest, Gradient Boost, Cat Boost, and Light Gradient Boosting Machine (LGBM). According to the study's results, LGBM performs better than the other four models since it has the highest recall and accuracy and the fewest False Negatives. The LGBM model's corresponding accuracy, precision, recall, f1 score, and AUC are 0.8789, 0.8978, 0.8553, 0.8758, and 0.9694. This demonstrates that, in comparison to traditional methods, machine learning optimization can produce notable advantages in churn risk classification. This study offers compelling proof that sophisticated machine learning modeling can revolutionize banking industry client retention management.

Exloratory Data Analysis Untuk Data Belanja Pelanggan dan Pendapatan Bisnis Widi Hastomo; Adhitio Satyo Bayangkari Karno; Sudjiran; Dodi Arif; Eka Sally Moreta
Infotekmesin Vol 13 No 2 (2022): Infotekmesin: Juli, 2022
Publisher : P3M Politeknik Negeri Cilacap

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35970/infotekmesin.v13i2.1547

A more quantifiable perspective is assuming the role of mechanistic management in an effort to enhance business based on its capacity to transform data into knowledge and insight. The industry has not completely supported its business strategy also with driven data. Using a transaction dataset taken from one of the Kaggle.com challenges, this experiment attempts to determine consumer spending patterns and Retail Fashion business revenues (H&M Personalized Fashion Recommendations). The results of the experiment are the number of transactions based on customer age, the most sales product and one-time purchased item, and the type of product that generates the highest and smallest income. The approach employed is EDA using the Python language. In order for businesses to generate analytical findings that provide future perspectives and to help identify the gap by delivering analytical results in the form of suggestions that can be perpetuated, the findings of this experiment are intended to support the capabilities of simulation. The challenge in this experiment is the abundance of datasets, which necessitates a suitable operating environment.

Title

Found 8 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 8 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 8 Documents
Search