Seismic　performance　prediction　of　slope　reinforcement　measures　is　an　essential　and　significant　problem　in　structure　design　and　monitoring　stage.　Enlightened　by　the　advancement　of　Recurrent　Neural　Network　(RNN)　in　geotechnical　engineering,　this　paper　develops　RNN　models　integrated　with　discrete　wavelet　transform　and　Dung　Beetle　Optimization　(DBO)　algorithms　for　predicting　the　dynamic　response　of　slope-pile-anchor　coupled　reinforcement　systems.　Centrifuge　shaking　table　tests　and　moving-steps　strategy　is　used　to　create　database　for　model　training.　Results　shows　that　DBO-Bidirectional　Long-Short　Term　Memory　(BiLSTM)　demonstrate　superiority　than　DBO-RNN,　DBO-LSTM　and　DBO-Gated　Recurrent　Unit　(GRU)　due　to　its　unique　bidirectional　learning　ability.　Moreover,　20　cases　involving　artificial　waves　with　an　average　PGA　of　0.1　g　were　predicted　using　the　DBO-BiLSTM　model　and　subsequently　compared　with　measured　data　for　validation.　The　results　indicate　that　the　proposed　DBO-BiLSTM　model　is　effective　for　time-series　prediction　of　seismic　responses　in　slope　reinforcement　structures.　This　model　provides　valuable　solutions　for　the　performance　design　and　health　monitoring　of　slope　engineering　structures,　particularly　in　situations　where　monitoring　data　are　limited.