Emotional　recognition　as　the　key　technology　in　the　field　of　emotion　computing　has　received　more　and　more　attentions　in　applications　such　as　human-computer　interaction,　medical-assisted　diagnosis　and　multimedia　intelligence　recommendation,　and　it　has　important　research　and　application　value.　EEG　recognition　based　on　EEG　is　a　commonly　used　and　effective　method　of　emotion　recognition.　More　and　more　scholars　are　concerned　with　the　emotional　recognition　of　specific　brain　regions　and　specific　frequency　bands　in　the　study　of　emotion　recognition　based　on　EEG.　They　will　be　based　on　the　distance　between　the　brain　electrode　and　the　principle　of　symmetry　divided　into　different　regions　to　extract　EEG　characteristics,　follow-up　identification　study.　But　this　way　ignores　the　emotional　correlation　and　difference　between　the　channels,　which　in　turn　affects　the　EEG　feature　representation　and　recognition　effect.　In　this　paper,　we　propose　a　channel　selection　method　based　on　the　maximum　correlation　minimum　redundancy　idea,　and　then　it　obtains　10　ideal　channels　for　the　futher　EEG　emotional　analysis.　©　2017　IEEE.