Sleep　apnea　syndrome　episodes　may　induce　high-risk　complications　such　as　pulmonary　hypertension,　cardiac　arrhythmia,　respiratory　failure,　and　hypertension.　It　is　of　great　significance　to　apply　neural　networks　for　efficient　automatic　diagnosis　of　sleep　apnea　syndrome.　We　propose　a　transfer　learning-based　classification　model　for　sleep　apnea　syndrome　using　ECG　signals　and　respiratory　signals,　which　results　in　a　91.26%　accuracy　in　recognizing　three　types　of　sleep　apnea　syndrome.