Computational Intelligence & Machine Learning (Coursework)
Coursework and assignments for the Kecerdasan Komputasional & Pembelajaran Mesin (MII226401) graduate course.
Things I have built and researched — web, mobile, games and applied AI/ML.
Coursework and assignments for the Kecerdasan Komputasional & Pembelajaran Mesin (MII226401) graduate course.
A modified build of the Eclipse Mosquitto MQTT broker (MQTT 5.0 / 3.1.1 / 3.1) with custom changes.
A Streamlit web app that classifies chest X-rays as tuberculosis-positive or negative using an SVM model. Digital image processing group project.
A Flutter mobile app for herbal medicine information, tutorials, recipes and a plant gallery, with an AI consultation feature backed by a Laravel API.
Source code for a hybrid facial emotion recognition model combining EmoNeXt and EfficientFER architectures.
Assignments and notebooks for the Pengolahan Citra Digital (MII226895) graduate course.
A Laravel backend and admin dashboard for a face-capture learning game, tracking student sessions and captured images with a REST API for the game client.
A Flask API that performs emotion detection from Facial Action Units (FAU), packaged with Docker.
CNN-based road condition classification with an Android GIS app for Simalungun Regency. My undergraduate thesis, also published as a journal article.
The Android client for E-Trash Clinic, a waste-management app where users submit waste, track prices and read recycling content.
The Laravel backend for E-Trash Clinic: a waste-management platform handling users, waste categories, pricing, articles and product results.
A Python server that runs anomaly/motif detection for the Fokus Teropong project using a TFLite autoencoder and Kernel Density Estimation.
An Android app for songket/weaving motif inference — capture an image, run on-device detection, and browse history of saved motifs.
The Laravel backend for the Pothole Tracker system, storing classified road segments and pothole locations reported by the Android app.
An Android app that captures road images, classifies road condition on-device, and tags detected potholes with GPS coordinates on a map.
An Android app ("Gen-Z Care") that estimates stress from facial analysis and a questionnaire, using an on-device TensorFlow model and sensor data.
Practice submissions (A, B, C) preparing for the TensorFlow Developer Certification, spanning regression, image, text and time-series problems.
A Java implementation of a genetic algorithm for nurse shift scheduling.
My personal Linux configuration: a tiling window manager setup with Hyprland, Waybar and related tools.
Modified firmware for the Huawei E5372 LTE mobile router, unlocking custom behaviour on the device.
An SMS gateway that sends and receives text messages through a Huawei E5372s LTE modem.
A Laravel web application (bankbook.my.id) with user authentication and account-style pages.
A cross-platform countdown timer app built with Flutter and Dart.
A TypeScript WhatsApp Multi-Device bot built on the multi-device (MD) protocol.
A Bukkit/Spigot Minecraft plugin that lets players store and withdraw their experience points.
A Bukkit/Spigot Minecraft plugin where zombies actively chase the player.
A pure-Python implementation that returns the array of integer points on a straight line between two coordinates, without NumPy.
A Node.js tool that splits MP4 files into HLS (HTTP Live Streaming) segments for adaptive web streaming.
A basic CRUD web service in Go using the Gin web framework and GORM ORM.
A Java Decision Support System implementing the Weighted Product (WP) method to rank alternatives against weighted criteria.
A Bukkit/Spigot Minecraft plugin that lets you customize the "player joined" and "player left" server messages.
A Bukkit/Spigot Minecraft plugin where lava chases the player, creating a rising-hazard survival challenge.