WIE3007 Data Mining Group Project

📊 AI-Enhanced Financial/Business Analytics

Course: WIE 3007 – Data Mining
Semester: 2025/2026 – Semester 1
Institution: Universiti Malaya, Faculty of Computer Science and Information Technology
Due: Week 13

---

👥 Team Members

Name	Matric No.	GitHub Username	Role
Ryan Chin Jian Hwa	23005233	@wrynaft	Model Evaluation
Liew Jin Sze	23005226	@jinsze	Data Modelling
Kueh Pang Lang	23005227	@pang-lang	EDA
Koay Khoon Lyn	23005235	@khoonlyn913	Dataset Simulation
Maxwell Jared Daniel	22002648	@oatmeal2211	Feature Engineering

---

📌 Project Overview

This project applies Data Mining and AI-enhanced analytics in the financial/business domain using Generative AI (GenAI), Large Language Models (LLMs), and Small Language Models (SLMs). The project encompasses dataset simulation, feature engineering, predictive modelling, and model interpretation with AI support.

Objectives

• Apply data-mining workflows in real financial/business contexts
• Integrate GenAI/LLMs/SLMs into data analysis
• Build and evaluate predictive models
• Collaborate effectively using professional GitHub practices

---

📋 Project Components

1️⃣ Dataset Simulation & Feature Engineering

• Simulate 1000+ financial/business-related records
• Use GenAI to create realistic numerical and textual patterns
• Apply LLMs/SLMs for feature extraction (sentiment analysis, risk categorization, customer segmentation)

2️⃣ Predictive Model Development

• Develop classification/regression models using:
  • Random Forest
  • Logistic Regression
  • XGBoost
  • Neural Networks
• Utilize AI tools for text-based feature engineering
• Compare results across at least two models

3️⃣ Model Evaluation & Interpretation

• Evaluate using appropriate metrics (Accuracy, F1-score, ROC-AUC, RMSE)
• Use LLMs to summarize findings and provide insights
• Interpret feature importance

4️⃣ Final Report

• 5–7 page comprehensive report
• Includes: objectives, dataset details, EDA, feature engineering, modelling, results, business insights
• AI usage disclosure and GitHub contribution summary

---