Update model/anomaly_detector.py

model/anomaly_detector.py

@@ -2,41 +2,39 @@ import numpy as np
 from sklearn.ensemble import IsolationForest
 from utils.preprocessing import extract_features
 
-# A placeholder Isolation Forest model
-# In production, load a pre-trained model using joblib or pickle
+# Dummy model with placeholder training
 model = IsolationForest(contamination=0.2, random_state=42)
-
-# Example training data (you should replace with real preprocessed dataset)
 example_training_data = np.array([
-    [5, 0, 30, 0],
-    [10, 1, 50, 1],
-    [2, 4, 10, 3],
-    [8, 0, 45, 0],
-    [3, 5, 12, 2]
+    [5, 0, 30, 0.1],
+    [10, 1, 50, 0.2],
+    [2, 2, 10, 0.3],
+    [8, 0, 45, 0.05],
+    [3, 1, 12, 0.4]
 ])
 model.fit(example_training_data)
 
 def detect_anomaly(agent_data):
     features = extract_features(agent_data)
-    prediction = model.predict([features])[0]  # -1 for anomaly
+    prediction = model.predict([features])[0]  # -1 = anomaly
     flagged = prediction == -1
 
-    reason = ""
-    if flagged:
-        if agent_data["missed_visits"] >= 5:
-            reason = "Missed appointments spike"
-        elif agent_data["weekly_calls"] < 4:
-            reason = "Low call frequency"
-        elif agent_data["lead_drop_rate"] > 0.6:
-            reason = "High lead drop rate"
-        else:
-            reason = "Unusual activity pattern"
+    # Override logic: manually flagging on rules
+    reasons = []
 
-    return {
-    "agent_id": str(agent_data.get("agent_id", "unknown")),
-    "weekly_calls": int(agent_data["weekly_calls"]),
-    "missed_visits": int(agent_data["missed_visits"]),
-    "flagged": bool(flagged),
-    "reason": str(reason if flagged else "Normal behavior")
-}
+    if agent_data["missed_visits"] >= 5:
+        reasons.append("Missed appointments spike")
+    if agent_data["weekly_calls"] < 3:
+        reasons.append("Low call frequency")
+    if agent_data["lead_drop_rate"] > 0.6:
+        reasons.append("High lead drop rate")
 
+    if reasons:
+        flagged = True
+
+    return {
+        "agent_id": str(agent_data.get("agent_id", "unknown")),
+        "weekly_calls": int(agent_data["weekly_calls"]),
+        "missed_visits": int(agent_data["missed_visits"]),
+        "flagged": bool(flagged),
+        "reason": ", ".join(reasons) if flagged else "Normal behavior"
+    }