Sistem Peringatan Dini Banjir Luapan Sungai
Dashboard Prediksi Curah Hujan Hulu Sumatera Utara (Conv1D-LSTM)
Pelatihan Jaringan Saraf Tiruan Conv1D-LSTM
Visualisasi profil optimasi model deep learning hulu secara real-time
Durasi Training Waktu Nyata
4.35s
Nilai Loss Epoch Akhir
| Epoch | Train Loss (MSE) | Val Loss (MSE) |
|---|---|---|
| 5 | 0.0095 | 0.0112 |
Arsitektur Jaringan:
Input → Conv1D(32, 3) → BatchNorm → MaxPooling1D(2) → LSTM(64) → Dropout(0.2) → Dense(12).