📈 Stock Price Prediction using LSTM | Future Interns ML Internship - Task 2

🔍 Task Overview

In Task 2, I tackled the problem of stock price prediction using Time Series Forecasting techniques. The project focuses on understanding market behavior and predicting future stock values — specifically for Apple Inc. (AAPL).

💡 What I Did

• 📥 Imported and cleaned the stock data (AAPL.csv)
• 📊 Performed exploratory data analysis (EDA) to visualize trends
• 🧠 Built a predictive deep learning model using LSTM (Long Short-Term Memory) networks
• 🧪 Evaluated model performance using RMSE and trend matching
• 📈 Visualized actual vs predicted stock prices

🧠 Key Learnings

• Time Series Decomposition and Feature Scaling
• Understanding windowing and sequence shaping for LSTM inputs
• The power of LSTM in capturing long-term dependencies in sequential data
• Techniques to reduce overfitting and improve model generalization

🚀 Model Performance

✅ The LSTM model produced highly accurate short-term predictions
📉 RMSE was kept low, and predicted values closely followed actual stock movements
📊 Visual alignment between the trend lines confirmed good predictive quality

🛠️ Tech Stack

• Language: Python
• Libraries:
  • pandas, numpy – Data preprocessing
  • matplotlib, seaborn – Data visualization
  • tensorflow.keras – LSTM model building
  • scikit-learn – Scaling and evaluation

🧪 How to Run

Clone the repository:

git clone https://github.com/Flash019/FUTURE_ML_02.git
cd FUTURE_ML_02
pip install -r requirements.txt
jupyter notebook FUTURE_ML_02.ipynb