Upload 5 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+dataset/card_transdata.csv filter=lfs diff=lfs merge=lfs -text

dataset/card_transdata.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7013c329bae9ef0ef32d65dbeb095694f0c7cd6c00ff74b2d0087fa1c67b8717
+size 76277977

models/fhe_files/client.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:322b3976b22d5ba9716880a6f414d5646709d0cf1604b708aaf2ce12194ab98f
+size 15633

models/fhe_files/server.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5fef601c70314eed11b39416aaa3cca9f3e0a8e63c8f53b97d3268797e3e56e
+size 9682

models/fhe_model.py

@@ -0,0 +1,57 @@
+"""
+Module for training and deploying an FHE-enabled
+Random Forest model using Concrete ML.
+"""
+
+import os
+import pandas as pd
+import joblib
+from sklearn.model_selection import train_test_split
+from sklearn.preprocessing import StandardScaler
+from concrete.ml.sklearn.rf import RandomForestClassifier
+from concrete.ml.deployment import FHEModelDev
+
+# Load the data (100,000 rows only)
+DATA_PATH = os.path.join(os.path.abspath(os.getcwd()), "dataset", "card_transdata.csv")
+df = pd.read_csv(DATA_PATH, nrows=100000)  # Limit to 100,000 rows
+
+# Check for missing values
+if df.isnull().sum().any():
+    df = df.dropna()
+
+# Handle class imbalance
+fraud = df[df["fraud"] == 1]
+non_fraud = df[df["fraud"] == 0].sample(n=len(fraud), random_state=42)
+balanced_df = pd.concat([fraud, non_fraud])
+
+# Separate features and target
+X = balanced_df.drop(columns=["fraud"])
+y = balanced_df["fraud"].astype(int)
+
+# Split into training and validation sets
+X_train, X_val, y_train, y_val = train_test_split(
+    X, y, test_size=0.2, random_state=42, stratify=y
+)
+
+# Preprocessing: scale the data
+scaler = StandardScaler()
+X_train_scaled = scaler.fit_transform(X_train)
+X_val_scaled = scaler.transform(X_val)
+
+# Save the scaler for later use
+SCALER_PATH = os.path.join(os.path.abspath(os.getcwd()), "models", "scaler.pkl")
+joblib.dump(scaler, SCALER_PATH)
+
+# Train the Random Forest model with Concrete ML
+model = RandomForestClassifier(n_estimators=100, random_state=42)
+model.fit(X_train_scaled, y_train)
+
+# Compile the model for homomorphic encryption
+model.compile(X_train_scaled)
+
+# Save the model and necessary files for client and server
+FHE_DIRECTORY = os.path.join(os.path.abspath(os.getcwd()), "models", "fhe_files")
+dev = FHEModelDev(path_dir=FHE_DIRECTORY, model=model)
+dev.save()
+
+print("Model trained, compiled, and saved.")

models/scaler.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d05215c687e429d53fd9e2ef3b461461e13b6f65ff532ca8d1ce55b89067de7a
+size 1231