Background:　Hepatocellular　carcinoma　(HCC)　is　a　common　malignant　tumor　with　high　morbidity　and　mortality　globally.　Compared　with　traditional　diagnostic　methods,　microRNAs　(miRNAs)　are　novel　biomarkers　with　higher　accuracy.　Objective:　We　aimed　to　identify　combinatorial　biomarkers　of　miRNAs　to　construct　a　classification　model　for　the　diagnosis　of　HCC.　Methods:　The　mature　miRNA　expression　profile　data　of　six　cancers　(liver,　lung,　gastric,　breast,　prostate,　and　colon)　were　retrieved　from　The　Cancer　Genome　Atlas　(TCGA)　and　Gene　Expression　Omnibus　(GEO)　database　with　accession　number　GSE36915,　GSE29250,　GSE99417,　GSE41970,　GSE64333　and　GSE35982.　The　messenger　RNA　(mRNA)　expression　profile　data　of　these　six　cancers　were　obtained　from　TCGA.　Three　R　software　packages,　student＇s　t-test,　and　a　normalized　fold-change　method　were　utilized　to　identify　HCC-specific　differentially　expressed　miRNAs　(DEMs).　Using　all　combinations　of　obtained　HCC-specific　DEMs　as　input　features,　we　constructed　a　classification　model　by　support　vector　machine　searching　for　the　optimal　combination.　Furthermore,　target　genes　prediction　was　conducted　on　the　miRWalk　2.0　website　to　obtain　differentially　expressed　mRNAs　(DEmRNAs),　and　KEGG　pathway　enrichment　was　analyzed　on　the　DAVID　website.　Results:　The　optimal　combination　consisted　of　four　miRNAs　(hsa-miR-130a-3p,　hsa-miR-450b-5p,　hsa-miR-136-5p,　and　hsa-miR-24-1-5p),　of　which　the　last　one　has　not　been　currently　reported　to　be　relevant　to　HCC.　The　target　genes　of　hsa-miR-24-1-5p　(CDC7,　ACACA,　CTNNA1,　and　NF2)　were　involved　in　the　cell　cycle,　AMPK　signaling　pathway,　Hippo　signaling　pathway,　and　insulin　signaling　pathway,　which　affect　the　proliferation,　metastasis,　and　apoptosis　of　cancer　cells.　Moreover,　the　area　under　the　receiver　operating　characteristic　curves　of　the　four　miRNAs　were　all　higher　than　0.85.　Conclusion:　These　results　suggest　that　the　miRNAs　combined　biomarkers　were　reliable　for　the　diagnosis　of　HCC.　Hsa-miR-24-1-5p　was　a　novel　biomarker　for　HCC　diagnosis　identified　in　this　study.