The Multivariate Statistical Analysis Of Chinese Rural Family Income Sources And Expenses Structure

This　paper　uses　several　statistical　methods　such　as　factor　analysis,　canonical　correlation　analysis　and　multivariate　linear　regression　to　establish　different　sources　of　Chinese　rural　family　income　and　expenses　with　the　aid　of　the　three　computer　programs　—MATLAB,　SPSS　and　SAS　,　and　therefore　analyzes　the　relationship　between　the　two　factors.　First　the　paper　classifies　31　provinces　or　municipalities　respectively　according　to　income　sources　and　expenses　by　using　factor　analysis.　This　is　done　objectively　and　reasonably　analyzes　the　actuality　and　its　causes,　as　well　as　giving　some　advice　to　improve　inhabitant　income　and　stimulate　expense　in　undeveloped　areas.　Secondly　the　paper　uses　average　and　power　less　square　regression　and　ridge　regression　to　establish　the　equation　of　basic　income　and　expense　variations.　Finally,　this　paper　indicates　the　relationship　between　the　two　variations　according　to　canonical　correlation　analysis.