Hi, I'm Umam!
I'm a Computer Science Student. While Iam immersed in the world of Data and AI research, I also enjoy exploring other specialities like alghorithms and software development.
ProjectsAbout Me
Get to know me!
Hi, my name is Moh Khoirul Umam Al Amin and I am a 5th semester computer science student at Binus University.
I am passionate about exploring machine learning. This include investigating the inner working of alghorithms.
Currently, I'm diving into the fascinating world of deep learning, exploring its concepts and applications. This journey of continuous learning allows me to stay up-to-date with the latest advancements in AI research and helps me understand the transformative potential of these technologies.🙂
I believe that I should never stop growing and that's what I strive to do, I am excited to see where my career takes me and am always open to new opportunities. 🙂
Projects
Parkinson Detection Using CNN
This project involves developing a neural network model to classify individuals as either having Parkinson's disease or being normal based on X-ray images. The objective is to leverage deep learning techniques to analyze medical imaging data and provide an accurate diagnostic
Machine Learning and Deep Learning Paper
This repository represents my self-study journey in machine learning and deep learning. I aim to implement concepts and architectures inspired by research papers, including Transformers, GANs, VAEs, and more. I strive to apply everything I've learned to deepen my understanding and gain hands-on experience in this exciting field
Indonesian Batik Image Classification
This project focuses on creating an intelligent solution for image classification by combining two powerful approaches: designing a custom Convolutional Neural Network (CNN) from scratch and leveraging the advantages of transfer learning with the pre-trained DenseNet121 model.
News Topic Classification Using LSTM
The project aims to develop a deep learning model using Long Short-Term Memory (LSTM) networks to classify news articles into predefined categories, such as politics, sports, technology, entertainment, and more. The model seeks to understand and analyze the context of news articles by learning from textual data.
Indonesian BankApps Sentiment Analysis
Conducting sentiment analysis on user reviews of three major Indonesian bank applications: BCA, BRI, and BNI. Utilizing the Sastrawi library, which provides powerful tools for text processing and stemming in the Indonesian language. Several classification algorithms, including Support Vector Machine (SVM), Random Forest, and XGBoost, are employed to categorize sentiments expressed in the reviews as positive, negative, or neutral.
BinusMaya Sentiment Analysis
As part of my academic research, I conducted a significant sentiment analysis project focused on BinusMaya, the primary learning management system used at Binus University. The study captured and analyzed feedback from over 600 students,providing valuable insights into user experience and satisfaction with the platform. The project implemented a dual-algorithm approach, utilizing both Naive Bayes and K-Nearest Neighbors (KNN) for classification.
Exploring Online Gambling in Indonesia
I developed an insightful data visualization project that provides a comprehensive analysis of online gambling patterns in Indonesia from 2017 through early 2024. The study offers a distinctive perspective on this sensitive social issue by visualizing key metrics including demographic distribution by age, transaction frequencies, and monetary values over a seven-year period.
Energy Consumption Forecasting
This project aims to forecast energy consumption patterns using historical data through time series analysis. By employing Facebook Prophet, the model captures seasonal trends and variations in energy usage. To enhance the forecasting accuracy, GridSearch is utilized to identify the optimal hyperparameters for the Prophet model.
Customer Segmentation
Transform raw customer data into actionable insights using advanced hierarchical agglomerative clustering. This intelligent segmentation system reveals natural customer groupings by analyzing multiple behavioral dimensions simultaneously, helping businesses deliver personalized experiences at scale.
