Computer Science (CO-SCIENCE)

Prediksi Risiko Alzheimer: Perbandingan Kinerja Algoritma Klasifikasi

2024-09-13T09:10:05+07:00

Alzheimer's disease is one of the most common forms of dementia characterized by a gradual decline in the ability to think, remember, and behave. The disease progresses slowly and usually begins with short-term memory loss, followed by difficulties in language, disorientation, and personality changes. Early detection of Alzheimer's is important to slow its progression. This study implements Knowledge Discovery in Database (KDD) with data pre-processing stages, SMOTE data balance, classification. The dataset from Kaggle contains 2,149 data 35 features. This study uses five types of algorithm models. with the five algorithms tested. Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, and K-Nearest Neighbor classification algorithms. Algorithm Implementation: The Random Forest algorithm is implemented during the classification process after data updates. This model is then assessed together with other algorithms using accuracy, precision, recall, F1-score, and ROC curve metrics to determine its effectiveness in detecting Alzheimer's. Meanwhile, the experiment was carried out by dividing the data into two parts, namely training data and test data of 70% and 30%. Research Results: The results showed that the Random Forest algorithm had the best performance with an Accuracy of around 0.98, Precision of 0.96, Recall of 1.00, and F1-score of 0.97, and an AUC value of around 0.99. This algorithm has proven to be superior to other algorithms in early detection of Alzheimer's disease.

Analisis Sentimen Cyberbullying Pada Komentar X Menggunakan Metode Naïve Bayes

2024-11-06T10:46:29+07:00

Advances in communication technology and social media have made it easier to access global information, but have also increased cases of cyberbullying on platforms such as X. The impact of cyberbullying can include physical and psychological disorders, such as increased loneliness, anxiety, depression, and decreased self-esteem. In addition, victims of cyberbullying may feel distress that can increase the risk of suicidal ideation. This research utilizes the Naïve Bayes method to effectively and efficiently classify cyberbullying-related comments. This classification model was developed to detect cyberbullying in comments on X, using the Naïve Bayes algorithm and a dataset from Kaggle consisting of 650 comments that contain cyberbullying characteristics and those that do not. This research includes several preprocessing steps such as tokenization, normalization, and stemming. The data was then divided into two parts: 80% for training data and 20% for testing data. The evaluation results show a model accuracy of 80.77%, precision 81.25%, recall 70.91%, and AUC 0.794. The innovation in this research lies in the use of 2 (two) stemming operators, namely stemming dictionary and stemming snowball, where the stemming dictionary uses a special file containing abbreviations or slang words, which are often used in comments on the word becomes its basic form. This model tends to be more accurate in classifying comments as non-bullying than bullying. Suggestions for improvement include exploring other preprocessing methods and algorithms, as well as using larger and more varied datasets.

Studi Perbandingan Algoritma Random Forest dan K-Nearest Neighbors (KNN) dalam Klasifikasi Gangguan Tidur

2024-11-06T10:42:58+07:00

Sleep disorders such as insomnia and sleep apnea can significantly affect quality of life and increase the risk of chronic diseases. Early identification and classification of sleep disorders are crucial in preventing further impacts. This study aims to compare the performance of the Random Forest and K-Nearest Neighbors (KNN) algorithms in classifying sleep disorders using the Sleep Health and Lifestyle Dataset from Kaggle, which contains health and lifestyle data relevant to sleep patterns. The Random Forest and KNN algorithms were applied to classify sleep disorders into the categories 'None', 'Sleep Apnea', and 'Insomnia'. Based on the study results, the Random Forest algorithm achieved an accuracy of 89.69%, with the best performance in the 'None' category, reaching a recall of 96.08%. Meanwhile, KNN achieved an accuracy of 87.02% with K=5. Although Random Forest demonstrated superior results, challenges were still found in detecting the 'Sleep Apnea' category, where recall only reached 74.55%, likely due to data imbalance. This study shows that the Random Forest algorithm is more effective in classifying sleep disorders compared to KNN. Future research steps include data balancing and exploring other algorithms such as XGBoost to improve the performance of sleep disorder detection.

Aplikasi Pencatatan Kalori Harian Berbasis Android Dengan Arsitektur MVVM

2024-12-05T15:49:18+07:00

An imbalance between calorie intake and expenditure is considered the main cause of obesity or being overweight. So by controlling your calorie intake in a balanced manner and according to your needs you will be able to prevent obesity. The daily calorie recording application can help someone record, control and obtain information on their calorie intake. This article discusses the development of this type of application, named Nutrizen, using the waterfall method, during the development process by the CH2-PS076 team in the Bangkit 2023 batch 2 program. The application created is an Android-based application created with the Kotlin programming language and MVVM architecture as a design pattern that is easy to learn and makes the code easy to understand and manage. Testing this application uses the usability test method with the System Usability Scale tool to determine the level of user acceptance of usability. The results obtained include a marginal level of usability acceptance, so improvements are needed so that this application can be more accepted and relied on by the wider community.

Analisis Sentimen Ulasan Pelanggan Menggunakan Algoritma Naive Bayes pada Aplikasi Gojek

2025-01-21T15:38:27+07:00

Transportation is a means that a person uses to move from one place to another. One mode of transportation that is popular among the public is online motorcycle taxis, such as Gojek. Gojek continues to innovate to meet customer needs more effectively, as well as expand the scope of its services. This research aims to identify the number of positive and negative sentiments in the user review dataset, evaluate the performance of the algorithm used, and measure the level of customer satisfaction with Gojek services. Analysis was carried out on 6,485 customer reviews, which resulted in 4,387 positive sentiments and 2,098 negative sentiments. The classification model used, namely Naive Bayes, shows an accuracy of 88.5%, precision of 88.1%, and recall of 89.0%. The results of this research indicate that the Naive Bayes method provides good performance in analyzing the sentiment of user reviews of Gojek services

Analisis Performa Model ResNet-50 Pada Diagnosis Pneumonia Balita Berdasarkan Citra Radiografi Thorax

2025-01-09T14:17:37+07:00

One of the most serious complications of ARI is pneumonia, where this disease causes sufferers to experience pain when breathing and limited oxygen intake. According to the World Health Organization (WHO), pneumonia is classified as a life-threatening disease due to the high mortality rate caused. To be able to diagnose this disease, patients usually undergo various medical examination methods, one of which is through chest radiography. However, the challenge in diagnosing pneumonia generally lies in the complexity and uncertainty in interpreting the results of these methods. Therefore, this study was conducted with the aim of building an image classification model based on the Chest radiography dataset from toddler patients using the ResNet-50 architecture, which is a variant of the Convolutional Neural Networks (CNN) algorithm. The combination of the two methods is applied to analyze and process images and obtain pattern recognition with high accuracy. The research methods used include the application of data augmentation, CNN architecture design, model training, and performance evaluation. The evaluation results show that the model has quite good performance with an accuracy of 85%, which indicates the model's ability to classify images with a fairly high level of accuracy, and has the potential to help the pneumonia diagnosis process more efficiently and accurately.

Analisa Komparasi Model Data Mining Algoritma C4.5, CHAID, dan Random Forest Untuk Penilaian Kelayakan Kredit

2024-11-06T11:09:41+07:00

Credit has now become a trend in society. The problem with credit is the improper history of credit card usage. The resulting impact can lead to bad credit. If customers fail to pay off debts that have been agreed upon with the bank, they can increase their credit risk. This study aims to conduct a comparative analysis of three data mining classification methods: the C4.5 algorithm, Chi-Squared Automatic Interaction Detection (CHAID), and Random Forest. The goal is to classify creditworthiness status. The researcher used 481 vehicle credit records with "bad" and "good" reviews. In this study, the independent variables used are dependent status, age, marital status, occupation, income, employment status, company status, last education, length of stay, house condition, and down payment. For creditworthiness assessment, the C4.5 model shows a truth accuracy rate of 91.90% with an area under the curve (AUC) value of 0.915. The CHAID model shows a truth accuracy rate of 63.83% with an AUC value of 0.661, and the Random Forest model shows a truth accuracy rate of 78.60% with an AUC value of 0.907. The evaluation results show that both the Random Forest and C4.5 algorithms have high accuracy rates and AUC values.

Customer Churn Prediction Pada Sektor Perbankan Dengan Model Logistic Regression dan Random Forest

2025-01-21T13:41:19+07:00

– Customer churn is a detrimental phenomenon in the banking sector because it can reduce revenue and increase the cost of acquiring new customers. This research aims to compare the performance of two models, Logistic Regression and Random Forest, to predict customer churn using datasets from Kaggle. The research process involves data preprocessing such as z-score normalization and dividing the dataset into training data (70%) and testing data (30%). The model was evaluated using a confusion matrix with Accuracy, precision, recall and F1-Score values. Logistic Regression achieved 76.85% Accuracy, 79% precision, 94% recall, and 86% F1-Score, showing quite good performance but susceptible to false positives. In contrast, Random Forest shows superior performance with 83.12% Accuracy, 84% precision, 96% recall, and 90% F1-Score. Random Forest is suitable for problems with high recall requirements because it is more reliable in detecting potential customer churn. To further improve model performance, it is recommended to perform hyperparameter optimization and feature importance analysis. This churn prediction model is expected to help banks reduce churn and increase customer retention.