Support　vector　machine　is　a　supervised　machine　learning　algorithm,　which　is　usually　solved　by　iterative　method.　Quantum　algorithms　have　significant　advantages　over　classical　algorithms　in　terms　of　speed　and　capacity.　In　this　work,　a　quantum　support　vector　machine　algorithm　based　on　gradient　descent　is　proposed.　Firstly,　in　the　training　stage,　quantum　gradient　descent　was　used　to　obtain　the　training　parameters,　which　reduces　the　time　complexity　of　a　single　iteration　from　the　classical　polynomial　level　to　the　logarithmic　level.　Then,　in　the　classification　stage,　we　built　a　quantum　classifier　that　can　classify　multiple　samples　at　the　same　time,　thereby　improving　the　efficiency　of　classification.　Finally,　the　simulation　experiments　proved　the　classification　ability　of　our　quantum　support　vector　machine.