The　impact　of　a　public　transport　fare　adjustment　on　the　travel　mode　choice　for　residents　is　studied　in　this　thesis.　Analyzing　the　macro　temporal　and　spatial　variation　of　public　transport　passenger　flow　characteristics　before　and　after　the　fare　adjustment,　it　was　found　that　there　was　a　great　impact　of　fare　adjustment　on　the　travel　of　residents.　Through　a　questionnaire　survey,　the　influence　level　of　the　fare　adjustment,　and　the　characteristics　of　travel　mode　transfer　for　commuters　and　non-commuters　were　different.　Based　on　the　temporal　and　spatial　feature　vectors　of　smart　card　data　for　public　transport　travelers,　the　commuters＇　classification　based　on　machine　learning　was　proposed.　Using　massive　public　transit　smart　card　data　of　about　13　million　every　day　in　Beijing,　including　rail,　bus　and　public　bicycles,　the　accurate　classification　of　commuters　and　non-commuters　was　achieved,　which　proved　that　the　classification　accuracy　reached　94.24%.　Based　on　the　accurate　classification　and　significance　analysis　on　the　temporal　and　spatial　trip　characteristics　of　different　types　of　public　transport　travelers,　the　effect　difference　of　the　public　transport　fare　adjustment　for　the　travel　frequency,　public　transport　travel　mode　choice　of　commuters　or　non-commuters　was　quantitatively　analyzed.　The　results　showed　that,　for　public　transport　commuters,　the　proportion　of　travelers　whose　travel　frequency　using　rail　and　bus　obviously　decreased　was　14.90%　and　25.47%,　respectively,　while　3.73%　of　commuters　transferred　from　rail　to　bus.　On　the　other　hand,　for　non-commuters,　the　proportion　of　travelers　whose　travel　frequency　using　rail　and　bus　obviously　decreased　was　21.32%　and　26.96%,　respectively.　The　decreasing　rate　of　rail　travel　frequency　for　non-commuters　was　greater　than　that　for　commuters,　while　the　decreasing　rate　of　bus　travel　frequency　for　two　types　of　travelers　was　basically　equal.　The　research　conclusion　provides　scientific　support　and　useful　reference　to　the　public　transit　network　planning,　operation　management　and　policy　implementation　impact　assessment.　©　2016　ASCE.