Project Overview

The goal of this project is to predict future stock prices based on historical stock data.
We used features like Open, High, Low, Close, and Volume to train a regression model and evaluate its prediction accuracy.

---

Dataset

• File: stock.csv
• Columns:
  • Date — Trading date
  • Open — Opening price
  • High — Highest price of the day
  • Low — Lowest price of the day
  • Close — Closing price
  • Volume — Trading volume

---

Tools & Libraries

• Python
• Pandas
• NumPy
• Matplotlib
• scikit-learn (Linear Regression, train-test split, metrics)

---

Data Preprocessing

• Converted Date column to datetime and sorted by date
• Handled missing values using forward fill
• Optional feature: Added Average price = (High + Low) / 2

---

Model Training

• Model: Linear Regression
• Features: Open, High, Low, Volume, Average
• Target: Close
• Train-Test Split: 80%-20% (time series, shuffle=False)

---

Model Evaluation

• Metrics:
  • Mean Squared Error (MSE)
  • R² Score
• Chart: Actual Close vs Predicted Close

---

Observations

• Linear Regression can predict general trend of stock prices
• Model performance can be improved with advanced models like LSTM or Random Forest Regression
• Key features influencing prediction: Open, High, Low, Volume, Average

---

How to Run

1. Place stock.csv in the project folder
2. Run stock_prediction.py in VS Code or any Python IDE
3. View terminal outputs and chart

---

Screenshots

• Terminal outputs: dataset info, MSE, R²
• Chart: Actual vs Predicted Close Price