My background

Hello! I'm Rendra Dwi Prasetyo as a Computer Science student at Bina Nusantara University. I have a strong passion for Data Science and the transformative power of data-driven insights. My journey in data science has been both challenging and rewarding, leading me to a variety of projects that highlight the endless potential of data in solving real-world problems.

Throughout my academic journey, I have actively explored data analysis, machine learning, and data visualization techniques, seeking to deepen my understanding of how data can drive better decision-making. I have built a solid technical background in Python, R, SQL and various machine learning tools and frameworks, which has helped me tackle tasks ranging from exploratory data analysis to predictive modeling.

My Contact

If you're interested in collaborating or connecting, click the icon below to reach out to me!

Education

Bina Nusantara Univeristy (Binus University)

Currently at semester 5

GPA : 3.80

This project is the final assignment for the Data Mining and Visualization course, focusing on analysis and interpretation related to mathematical and computational decision-making using the R programming language. This project aims to analyze a dataset on the distribution of poverty and social assistance in Bandung City in 2022 using the R language and the Rpubs platform. By using descriptive statistical methods, preprocessing, data grouping and data visualization, this report will provide a comprehensive picture of the distribution of social assistance and poverty in the city of Bandung in 2022.

This project aims to build a Convolutional Neural Network (CNN) model that can distinguish between normal lung images and those affected by pneumonia. The model uses a dataset of lung X-ray images to detect health conditions, focusing on binary classification (lung normal vs lung affected by pneumonia). The ultimate goal is to develop the best model that assists in the early detection of pneumonia with high accuracy. On this project, the best model I achieved reached an accuracy of 89%.

This project aims to explore and maximize the performance of text classification models through text mining and hyperparameter tuning. I use two main machine learning algorithms, Support Vector Machine (SVM) and Random Forest, then for text representation I use the TF-IDF and Bag of Words (BoW) vectorization methods. The best accuracy results I achieved. With a combination of TF-IDF and linear kernel SVM, I achieved 96% accuracy in classifying text data. The project also provides detailed comparisons between various model combinations to outline their advantages and disadvantages. Through this project, I demonstrated my ability to carry out text exploration, model optimization, and performance analysis with systematic and in-depth methods.

My project aims to explore and compare the effectiveness of two machine learning algorithms : Support Vector Machines (SVM) and BERT (Bidirectional Encoder Representations from Transformers) for conducting sentiment analysis on user reviews of the Gojek application. Our experimental results show that BERT outperforms SVM due to its more complex and in-depth approach to text processing.

On this project my team and I developing a Database Management System (DBMS) using SQL language to streamline the new student admission process for a fictional university, Binus (Bina Negeri University), located in Indonesia. The growing number of applicants and the complexity of the admission process necessitated a computerized solution to manage applicant data efficiently, verify documents, and support decision-making.

My team and I developed "Lifevel Up" an Android application created with Flutter which has features in the form of presenting material, tests and quizzes for students. This project was developed based on our goal to help people who are struggling to improve themselves due to traditional and ineffective methods. My main role in this project was as a project lead to help my team in various areas. I always give my best when working on a project. I have contributed to every part of this project from brainstorming, planning, presentation, and application development.

I developed a text classification application that utilizes machine learning models to categorize text into two sentiment classes: positive or negative. The application allows users to choose from four different models: Support Vector Machine (SVM), Naive Bayes, Logistic Regression, and Random Forest. This project spans from dataset analysis to model training and final deployment, providing a reliable tool to determine sentiment polarity in text.

On this project, I developed a machine learning model to recommend books based on user preferences. The system's core is a K-Nearest Neighbors (KNN) model, which identifies and recommends books similar to a user-selected title. The model processes a large dataset of book features and uses distance-based metrics to compute the most relevant recommendations.

In this project, I developed a Generative Adversarial Network (GAN) to generate synthetic images, demonstrating proficiency in deep learning and generative modeling. The GAN consists of a generator that creates realistic images and a discriminator that distinguishes real images from synthetic ones. The two networks are trained adversarially to improve each other's performance. Initially, I utilized the MNIST dataset to generate handwritten digit images, achieving realistic outputs. This project showcases my understanding of neural network design, adversarial training, and model convergence.

This project investigates the performance of Transformer-based models, specifically BERT, compared to a traditional machine learning approach (TF-IDF + SVM) for sentiment classification. The goal is to analyze the strengths and weaknesses of each method and understand their behavior on nuanced text data.