Big　data　and　Internet　of　Things　are　playing　indispensable　roles　in　human’s　daily　life.　In　the　era　of　information,　all　things　are　linked　and　enhanced　by　the　Internet,　from　original　virtual　data　to　daily　essentials,　known　as　Internet　of　Things,　from　national　strategy　to　urban　development,　known　as　smart　city.　Abundant　digital　sources　bred　by　a　variety　of　applications　are　not　only　a　feature　but　also　a　challenge　of　big　data　analysis.　This　chapter　focuses　on　large-scale　data　processing　methods　suitable　for　big　data　analysis　scenarios　in　smart　city,　mainly　the　large-scale　machine　learning.　The　chapter　presents　six　prevalent　directions　for　classification　distributed　optimization　evolving　from　standalone　mode　working　to　cluster　mode.　The　classical　classification　algorithms　are　intrinsically　sequential　violating　the　parallel　framework;　therefore,　several　types　of　parallelization　approaches　are　proposed　to　adapt　stale　mechanisms　to　advanced　distributed　fashions.　These　approaches　include　a　variety　of　improvements　facing　distinct　issues.　In　addition,　this　chapter　discusses　the　basic　statistical　learning　problems,　in　which　classification　conforms　to　a　generic　paradigm,　and　details　theoretical　analysis　for　each　method.　Additionally,　a　brief　case　study　is　given　to　gain　deep　insight　of　practical　applications.　©　2018　by　Taylor　&　Francis　Group,　LLC.