The　study　on　using　multiscale　entropy　to　extract　emotional　EEG　features　was　conducted,　and　considering　that　the　traditional　feature　extraction　algorithm　using　multi-scale　entropy　may　result　in　loss　of　important　information　during　the　coarsening　process　and　the　scale　selection　problem:　small-scale　causes　no　significant　feature　while　big-scale　causes　excessive　computation,　an　improved　extraction　algorithm　based　on　multi-scale　entropy　was　put　forward.　The　improved　algorithm　determines　scales　according　to　the　number　of　intrinsic　mode　functions　of　the　adaptive　multiscale　entropy,　and　uses　the　adaptive　method　to　perform　the　binay-state　processing　of　the　EEG　signal　to　highlight　the　EEG　signal＇s　small　changes　thus　the　characteristics　of　the　EEG　signal　can　be　fully　tapped,　and　the　complexity　of　the　algorithm　can　be　reduced.　Based　on　the　optimized　SVM　(support　vector　machine)　classifier,　the　emotional　EEG　recognition　was　achieved　by　using　the　international　Deap　database　for　emotion　analysis,　and　the　performance　of　the　improved　algorithm　was　compared　with　traditional　algorithms.　The　results　indicated　that　the　classification　accuracy　of　the　improved　algorithm　was　higher　12.33%　compared　with　the　traditional　multiscale　entropy　algorithm,　and　higher　7.27%　compared　with　the　adaptive　multiscale　entropy　algorithm,　showing　that　the　improved　algorithm　is　effective　for　extracting　EEG　features.　©　2015,　Inst.　of　Scientific　and　Technical　Information　of　China.　All　right　reserved.