Add SDV privacy classification project

class_project/data605/Spring2026/projects/UmdTask391_DATA605_Spring2026_Synthetic_Data_Vault/Dockerfile

@@ -0,0 +1,10 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY . .
+
+CMD ["python", "run_project_check.py"]

class_project/data605/Spring2026/projects/UmdTask391_DATA605_Spring2026_Synthetic_Data_Vault/README.md

@@ -0,0 +1,89 @@
+# DATA605 Spring 2026: Synthetic Data Vault Privacy Classification Project
+
+## Project Overview
+
+This project explores the use of Synthetic Data Vault (SDV) for privacy-preserving synthetic data generation.
+
+The main goal is to generate a synthetic version of a real-world dataset and test whether the synthetic data can still support a machine learning classification task. I used the Adult Income dataset, generated synthetic training data using SDV, and compared models trained on real data versus synthetic data.
+
+## Dataset
+
+The project uses the Adult Income dataset. The dataset contains demographic and employment-related features such as age, education, occupation, work hours, and income class.
+
+The target variable is:
+
+- `<=50K`
+- `>50K`
+
+This is a binary classification problem.
+
+I selected this dataset because it contains both numerical and categorical features, making it a good example for testing synthetic tabular data generation.
+
+## Tools and Libraries
+
+Main tools used in this project:
+
+- Python
+- Pandas
+- NumPy
+- Matplotlib
+- Seaborn
+- Scikit-learn
+- Synthetic Data Vault
+- SDMetrics
+- Jupyter Notebook
+
+## Project Workflow
+
+The project follows these steps:
+
+1. Load the Adult Income dataset.
+2. Clean missing and inconsistent values.
+3. Explore the dataset using summary statistics and visualizations.
+4. Split the real data into training and testing sets.
+5. Fit an SDV Gaussian Copula synthesizer on the real training data.
+6. Generate a synthetic training dataset.
+7. Evaluate synthetic data quality using SDMetrics.
+8. Train classification models on real and synthetic data.
+9. Evaluate all models on the same real test set.
+10. Compare accuracy, precision, recall, and F1-score.
+11. Run a small hyperparameter tuning experiment.
+12. Interpret results, limitations, and conclusions.
+
+## Main Experiment
+
+The most important experiment is:
+
+> Train a machine learning model on synthetic data and test it on real unseen test data.
+
+This checks whether the synthetic data preserves useful predictive patterns from the original dataset.
+
+## Files
+
+```text
+notebooks/Synthetic_Data_Vault.ipynb
+    Main notebook containing the full SDV workflow, EDA, model training, evaluation, and interpretation.
+
+outputs/model_comparison_results.csv
+    Initial model comparison results.
+
+outputs/final_model_comparison_results.csv
+    Final results including the hyperparameter tuning experiment.
+
+requirements.txt
+    Python libraries required to run the project.
+
+synthetic_data_vault_utils.py
+    Utility file reserved for helper functions.
+
+Dockerfile
+docker_build.sh
+docker_run.sh
+    Docker-related project files.
+
+## Additional Tutorial Notebook
+
+I also included a shorter API-focused notebook:
+
+```text
+notebooks/synthetic_data_vault.API.ipynb

class_project/data605/Spring2026/projects/UmdTask391_DATA605_Spring2026_Synthetic_Data_Vault/docker_build.sh

@@ -0,0 +1,10 @@
+#!/usr/bin/env bash
+
+set -e
+
+IMAGE_NAME="sdv-privacy-classification"
+
+echo "Building Docker image: ${IMAGE_NAME}"
+docker build -t ${IMAGE_NAME} .
+
+echo "Docker image built successfully."

class_project/data605/Spring2026/projects/UmdTask391_DATA605_Spring2026_Synthetic_Data_Vault/docker_run.sh

@@ -0,0 +1,8 @@
+#!/usr/bin/env bash
+
+set -e
+
+IMAGE_NAME="sdv-privacy-classification"
+
+echo "Running Docker image: ${IMAGE_NAME}"
+docker run --rm ${IMAGE_NAME}