Add Deep Learning (LSTM) + Quick Wins automation features

Major improvements:
- Deep Learning integration with PyTorch LSTM (Bidirectional, 128→64 units)
- Hybrid predictor: 40% RandomForest + 60% Deep Learning
- LaunchAgent for automatic weekly tip generation (Tue/Fri 21:00)
- Health-Check system with auto-recovery and Telegram alerts
- Model caching and intelligent retraining logic
- Updated CSV data and generated tips
- Performance reports for recent draws

Technical details:
- PyTorch used instead of TensorFlow (Python 3.14 compatibility)
- Apple Silicon MPS acceleration support
- Sequence learning with 20-draw history
- Early stopping and learning rate scheduling

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

data/learning_log.json

@@ -43,6 +43,58 @@
         "main": 1.2,
         "sz_rate": 0.1
       }
+    },
+    {
+      "timestamp": "2025-12-22T21:00:50.378355",
+      "draw_date": "2025-12-20",
+      "evaluation": {
+        "main": 3,
+        "sz": false,
+        "tip": 2
+      },
+      "avg_matches": {
+        "main": 0.7,
+        "sz_rate": 0.1
+      }
+    },
+    {
+      "timestamp": "2025-12-25T21:00:35.769591",
+      "draw_date": "2025-12-24",
+      "evaluation": {
+        "main": 2,
+        "sz": false,
+        "tip": 2
+      },
+      "avg_matches": {
+        "main": 0.8,
+        "sz_rate": 0.1
+      }
+    },
+    {
+      "timestamp": "2025-12-29T21:00:36.294861",
+      "draw_date": "2025-12-27",
+      "evaluation": {
+        "main": 2,
+        "sz": false,
+        "tip": 1
+      },
+      "avg_matches": {
+        "main": 1.2,
+        "sz_rate": 0.1
+      }
+    },
+    {
+      "timestamp": "2026-01-05T11:56:39.905126",
+      "draw_date": "2026-01-03",
+      "evaluation": {
+        "main": 2,
+        "sz": false,
+        "tip": 1
+      },
+      "avg_matches": {
+        "main": 1.5,
+        "sz_rate": 0.0
+      }
     }
   ]
 }