Main project: Stock Market Trend Analysis — forecast next-day price movement for the S&P 500 and three tech stocks (AAPL, AMZN, NVDA), comparing classical (ARIMA/GARCH), tree (LightGBM), and deep (LSTM) models. Honest result: no usable out-of-sample price skill (efficient-market ceiling), but volatility is predictable.
Educational only — not investment advice.
# Get the project files (the project is a subfolder, so we move it into place)
!rm -rf /content/dsp /content/Stock-Market-Trend-Analysis
!git clone https://github.com/Abdooo2235/Data-Science-Project.git /content/dsp
!mv /content/dsp/Stock-Market-Trend-Analysis /content/Stock-Market-Trend-Analysis
!pip install -q -r /content/Stock-Market-Trend-Analysis/requirements.txt
print("Ready. Now open a notebook from notebooks/ and Run all.")File → Open notebook → GitHub → search Abdooo2235/Data-Science-Project → pick one:
| Notebook | What it does |
|---|---|
notebooks/01_data_collection_preprocessing.ipynb |
Fetch + clean + engineer features (builds the data) |
notebooks/02_eda.ipynb |
Exploratory data analysis (18 figures) |
notebooks/03_model_building.ipynb |
ARIMA + GARCH + LightGBM + LSTM |
notebooks/04_evaluation_presentation.ipynb |
Final evaluation |
notebooks/04_volatility_modeling.ipynb |
Volatility forecasting |
notebooks/05_volatility_regime.ipynb |
Calm-vs-stormy regime classifier |
That's it.
Notes
- The processed data ships with the repo, so 02/03/04/05 run on their own — you don't have to run 01 first.
- Repo is public — no login or token needed.
- For the LSTM in notebook 03, pick a GPU:
Runtime → Change runtime type → T4 GPU.
git clone https://github.com/Abdooo2235/Data-Science-Project.git
cd Data-Science-Project/Stock-Market-Trend-Analysis
py -3.10 -m venv .venv && .venv\Scripts\activate # Windows
python -m pip install -r requirements.txt
python notebooks/01_data_collection_preprocessing.py # then 02, 03, ...Full architecture, decisions, and milestone reports: Stock-Market-Trend-Analysis/README.md.