About
Final Year AI Engineering Student at UiTM
Projects
This project aims to build a machine-learning model to detect suicidal ideation and related distress in tweets posted on X (previously known as Twitter). With the increasing use of social media platforms for expressing emotions, detecting suicidal content has become crucial in identifying individuals at risk and providing timely intervention.
The project implements a Natural Language Processing (NLP)-based classification model to analyze tweet texts and categorize them as either “Suicidal” or “Non-Suicidal.” Additionally, it performs sentiment analysis to evaluate the emotional tone of the tweets further, providing insights into the level of distress by classifying tweets as positive, negative, or neutral. This dual-layered approach not only helps in detecting suicidal tendencies but also assists in understanding the underlying emotional states of users.
Key features of the project include:
-
Preprocessing of tweet texts to clean and normalize the data.
-
Feature extraction using popular NLP techniques (e.g., TF-IDF, word embeddings).
-
Training a machine-learning model for suicidal ideation detection.
-
Sentiment analysis to categorize the emotional tone of tweets.
-
Evaluation of model performance using appropriate metrics like accuracy, precision, recall, and F1-score.
This repository contains the code and resources for implementing both the suicidal tweet detection and sentiment analysis tasks, aimed at enhancing mental health support and timely interventions on social media.
Model Accuracy
-
Logistic Regression = 0.87
-
Decision Tree = 0.85
-
Random Forest = 0.89
-
Naive Bayes = 0.86
API with Flask
This project includes a Flask-based API that serves predictions for suicidal ideation based on user-submitted tweets. The API endpoints allow users to send tweets and receive predictions in real-time.
HeartCare: Heart Disease Prediction System using Machine Learning Algorithms
Heart disease is a leading cause of mortality worldwide and, early detection and effective management are critical for improving patient outcomes. Despite the advancements in medical technology, there is a need for innovative tools to assist in continuous monitoring and prediction of heart disease. This research aims to develop a machine learning-based prediction model that can help detect and assess the risk of heart disease in individuals. By analyzing a comprehensive dataset of patient information, including demographic data, medical history, lifestyle factors, and clinical measurements, the model will provide healthcare professionals with a tool for early identification of at-risk individuals.
Awards
We participated in IREx 2024 with our group project "MyPredictC": Covid-19 Prediction that showcases our Covid-19 dashboard that can identify high-risk areas, analyze cluster trends and durations, and closely monitor total cases within each cluster. Our group managed to get the silver award in the IREx 2024.
Participated in a student design competition to showcase our mobile application design and managed to get a silver award
Work Experience
Responsible for frontend development for Kitahack 2024's website with other team members
-
Proficiently managed time to handle multiple tasks, including creating visually engaging content using DaVinci Resolve, Adobe Premiere Pro, Canva, and Adobe Photoshop during my tenure at Dinamiq Agency.
-
Collaborated effectively within a team environment, contributing to web design projects using Elementor and WordPress, showcasing strong teamwork and communication skills.
-
Demonstrated versatility by assisting in IT tasks for Dinamiq Agency employees, showcasing technical support capabilities.
-
Created impactful posters and visual materials, highlighting a blend of creativity and design expertise in diverse projects undertaken during the internship.
-
Applied JavaScript, HTML, and CSS to enhance website functionality and aesthetics, ensuring a seamless user experience and further showcasing technical prowess and attention to detail.