feat(ml): replace Autoencoder with RealNVP Normalizing Flow and add SessionTransformer embeddings
Replace TrafficAutoEncoder (MSE reconstruction scoring) with TrafficNormalizingFlow (RealNVP via FrEIA, 4 affine coupling blocks, anomaly score = -log p(x)) for mathematically rigorous density estimation. Add SessionTransformer module producing 32-dimensional sequence embeddings from raw HTTP request sequences (path, method, timing) via a lightweight TransformerEncoder, replacing path_transition_entropy and cadence_cv features. Update thesis documentation sections 2.4.2b and 3.8 accordingly. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
@ -430,8 +430,7 @@ Chaque anomalie reçoit un `campaign_id` (−1 = pas de cluster).
|
||||
|
||||
| Feature | Description |
|
||||
|---------|-------------|
|
||||
| `path_transition_entropy` | Entropie des transitions de chemins (§5.1) |
|
||||
| `cadence_cv` | Coefficient de variation de la cadence de requêtes (§5.3) |
|
||||
| `seq_emb_0`..`seq_emb_31` | Embeddings séquentiels via Transformer (§5.2, remplace path_transition_entropy + cadence_cv) |
|
||||
| `burst_ratio` | Fraction de requêtes en rafale (§5.3) |
|
||||
| `pause_ratio` | Fraction de pauses longues (§5.3) |
|
||||
| `lag1_autocorrelation` | Autocorrélation lag-1 des inter-arrivées (§5.3) |
|
||||
|
||||
Reference in New Issue
Block a user