Fendy Hendriyanto

SIPEKA (Sistem Prediksi Karyawan Resign)

SIPEKA (Sistem Prediksi Karyawan Resign) is an initial application launched at the end of this month to address the concerns of a company's HR team, who wanted to monitor employee turnover to determine whether the risk of employee resignation was high or stable. This application was built using a machine learning model based on the ensemble algorithm Random Forest, combined with SMOTE (Synthetic Minority Over-sampling Technique) preprocessing, using an imbalanced dataset. The model evaluation achieved excellent results with 98% accuracy, 97% F1-Score, 97% precision, 95% recall, and 98% ROC-AUC score.

NULISING (Nulis Komering)

NULISING (Nulis Komering) is a web application platform for recognizing Komering script using Optical Character Recognition (OCR), dedicated to recognizing, translating, and teaching the Komering script. The platform will allow users to scan script text (from both ancient manuscripts and handwritten text) and instantly obtain its Latin transliteration and meaning. The expected impact is the revitalization of the Komering script through AI-based technology.

Automation Scheduling For Shifting Employee Using Genetic Algorithm With DEAP

Application for automation scheduling for shifting employees using DEAP framework with the Flask for deployment.

Customer Retail Prediction (Case: Turkey Customer)

Dashboard Analysis and performing predictive analysis for customer retail in Turkiye country with Ensemble Technique based.

Padiku (Dashboard & Rice Prediction)

Creating a dashboard and performing predictive analysis for rice production on Sumatra Island using a regression-based algorithm.

Cirebon Mask Identification Using Combined Feature Texture and CNN Algorithm

A project with development and improvement from the previous method with a deep learning based for improve accuracy and predicting model with new data.

Dashboard Netflix Analysis

A dashboard with case Netfix data users and movies analysis using Tableau Public.

AI Chatbot Kampus Merdeka using LSTM

This project is about the AI Chatbot Kampus Merdeka to help students or Indonesian people know about the Kampus Merdeka program from KEMENDIKBUDRISTEK (Kementerian Pendidikan, Kebudayaan, Riset Dan Teknologi) or Ministry of Education, Culture, Research and Technology of the Republic of Indonesia.

Scale Invariant Feature Transform for Cirebon Mask Classification

The vast diversity of art in Indonesia generates many interests both domestically and internationally. One of the prominent cultures is Cirebon Mask. There are five types of Cirebon masks: Panji, Samba, Rumyang, Tumenggung, and Klana. In this research project, Cirebon masks are classified using digital images processing techniques using Scale-Invariant Feature Transform, while K-Nearest Neighbour, Support Vector Machines, and Random Forest as the classifier with MATLAB programming language.

Spam Text Classification Using Recurrent Neural Network (RNN)

This project is about Artificial Intelligence Domain Natural Language Processing with the title Spam Text Classification Using Recurrent Neural Network. Text Classification is a part of NLP for the identification of text or document. Spam text classification is useful for seeing which messages contain spam or not in email messages. Hopefully, this project repository can help to be used as reference material.

Rock, Paper, and Scissors Classification Using CNN and RNN-LSTM Algorithm

This project is part of the final Dicoding Indonesia submission. This project too, uses a deep learning algorithm, namely Convolution Neural Network (CNN) and Recurent Neural Network (RNN) for image classification with rock, paper, and scissors hand datasets.

Cirebon Mask Classification using Robust K-Nearest Neighbour

Indonesian culture captivates people both domestically and internationally. One of the prominent cultures is Cirebon Mask. There are five types of Cirebon masks: Panji, Samba, Rumyang, Tumenggung, and Kelana. In this work, we present texture features using the first-order statistical feature with outlier detection using z-score while K-Nearest Neighbour acted as the classifier. From this experiment, our proposed methods increase the accuracy up to 10% compared to the basic method.

Cirebon Mask Feature Extraction Using Combined Segmented Fractal Texture Analysis and Gray Level Co-Occurrence Matrix Feature

A final thesis project related to AI for Humanities, combining segmented fractal texture analysis and Gray Level Co-Occurrence Matrix features to represent classical mask textures for machine learning classification in performing arts data.

• This topic was presented in a series of events at the Python Conference Indonesia in 2023
• One of the AI for Humanities projects is carried out from various cultural data that does not receive enough attention, such as data on Classical Masks for performing arts, especially Cirebon Masks.
• The third research topic is related to the case of mask data with digital image processing using machine learning models. Previously, it was related to the Balinese Mask with the CNN algorithm and the Yogyakarta Dagelan Mask with the Eigenfaces algorithm.