Data Analytics and Intelligence

Overview

This project was completed as part of the I304 – Data Analytics and Intelligence course.
It applies data analytics and machine learning techniques using Python to solve real-world business intelligence problems.

The assignment consists of four major tasks:

1. Linear Regression – Stock price prediction using Yahoo Finance API
2. Data Clustering – Credit card dataset clustering using K-Means
3. Data Classification – Breast Cancer Wisconsin dataset classification
4. Principal Component Analysis (PCA) – Dimensionality reduction with classification

---

Technologies Used

• Python 3.x
• Pandas, NumPy – Data manipulation
• Matplotlib, Seaborn – Data visualization
• scikit-learn – Machine learning (Regression, Classification, Clustering, PCA)
• yfinance – Stock market data retrieval

---

Tasks Breakdown

Task 1: Linear Regression – Stock Price Prediction

• Retrieved stock data using Yahoo Finance API
• Performed exploratory data analysis on stock features
• Built a multivariate linear regression model to predict stock closing prices
• Evaluated model performance with R² score
• Tested predictions on custom input values

Task 2: Data Clustering – Credit Card Dataset

• Applied K-Means clustering (clusters = 3 → 15)
• Visualized results with different color-coded clusters
• Determined the most suitable number of clusters

Task 3: Data Classification – Breast Cancer Dataset

• Preprocessed and scaled the dataset
• Trained two classifiers:
  • Logistic Regression
  • K-Nearest Neighbours (KNN)
• Compared accuracy results between models

Task 4: Principal Component Analysis (PCA)

• Performed PCA with n_components=2 on Wine dataset
• Reduced dimensionality while preserving key variance
• Trained a Logistic Regression classifier on reduced data
• Visualized results with a scatter plot

---

How to Run

1. Clone this repository:

   git clone https://github.com/yourusername/data-analytics-assessment1.git
   cd data-analytics-assessment1

2. Install dependencies:

   pip install -r requirements.txt

3. Run any task script, e.g.:

   python Task1_LinearRegression/linear_regression_stock.py

---

Results Summary

• Regression: Achieved meaningful stock price predictions with acceptable R² scores.
• Clustering: K-Means formed well-separated customer groups in the credit card dataset.
• Classification: Both Logistic Regression and KNN achieved strong accuracy; Logistic Regression performed slightly better.
• PCA: Successfully reduced data to 2D while maintaining classification effectiveness.