Titanic Dataset

This project uses the Titanic training dataset from Kaggle after extrapolation to 1M dataset size.

Download the dataset and place it in the data/ directory before running the agent.

---

Running the Agent

You can run the agent in two different modes:

1. GPU-Accelerated Mode (NVIDIA cuML + cuDF)

Leverages NVIDIA's RAPIDS libraries for faster data processing and model training.

python -m cudf.pandas -m cuml.accel run.py

---

2. CPU Mode (scikit-learn + pandas)

Uses standard pandas and scikit-learn for data processing and modeling.

python run.py

---

Note:

• Ensure you have the appropriate dependencies installed for each mode.
• GPU mode requires a supported NVIDIA GPU and the RAPIDS ecosystem installed.