Create financial_ai_model.py

financial_ai_model.py

@@ -0,0 +1,76 @@
+import numpy as np
+import pandas as pd
+import tensorflow as tf
+from sklearn.ensemble import IsolationForest
+from sklearn.preprocessing import MinMaxScaler
+from tensorflow.keras.models import Sequential
+from tensorflow.keras.layers import LSTM, Dense, Dropout
+
+# Load data transaksi (dummy data)
+def load_transaction_data():
+    data = {
+        "timestamp": pd.date_range(start="2025-01-01", periods=500, freq="H"),
+        "amount": np.random.uniform(1, 1000, 500),
+        "fee": np.random.uniform(0.1, 5, 500),
+        "balance": np.cumsum(np.random.uniform(-50, 50, 500))
+    }
+    df = pd.DataFrame(data)
+    return df
+
+# **1. Deteksi Anomali Transaksi dengan Isolation Forest**
+def detect_anomalies(df):
+    model = IsolationForest(n_estimators=100, contamination=0.05, random_state=42)
+    df["anomaly"] = model.fit_predict(df[["amount", "fee", "balance"]])
+    return df
+
+# **2. Prediksi Tren Keuangan dengan LSTM**
+def prepare_lstm_data(df):
+    scaler = MinMaxScaler()
+    df_scaled = scaler.fit_transform(df[["balance"]])
+
+    X, y = [], []
+    for i in range(len(df_scaled) - 10):
+        X.append(df_scaled[i:i+10])
+        y.append(df_scaled[i+10])
+
+    return np.array(X), np.array(y), scaler
+
+def build_lstm_model():
+    model = Sequential([
+        LSTM(50, activation='relu', return_sequences=True, input_shape=(10, 1)),
+        Dropout(0.2),
+        LSTM(50, activation='relu'),
+        Dropout(0.2),
+        Dense(1)
+    ])
+    model.compile(optimizer='adam', loss='mse')
+    return model
+
+# **3. Rekomendasi Strategi Transaksi**
+def transaction_recommendation(df):
+    avg_fee = df["fee"].mean()
+    if df["balance"].iloc[-1] < 100:
+        return "Saldo rendah! Kurangi transaksi besar."
+    elif avg_fee > 2:
+        return "Biaya transaksi tinggi. Gunakan strategi batch processing."
+    else:
+        return "Transaksi dalam batas normal."
+
+# **Menjalankan Model**
+df = load_transaction_data()
+df = detect_anomalies(df)
+
+X, y, scaler = prepare_lstm_data(df)
+model = build_lstm_model()
+
+# **Latih Model**
+model.fit(X, y, epochs=10, batch_size=16, verbose=1)
+
+# **Prediksi Tren Masa Depan**
+future_balance = model.predict(X[-1].reshape(1, 10, 1))
+future_balance = scaler.inverse_transform(future_balance.reshape(-1, 1))
+
+# **Output Analisis**
+print("\n🔍 Anomali yang terdeteksi:", df[df["anomaly"] == -1][["timestamp", "amount", "fee"]])
+print("\n📈 Prediksi saldo keuangan selanjutnya:", future_balance[0][0])
+print("\n✅ Rekomendasi transaksi:", transaction_recommendation(df))