The　global　primary　health　concern　of　skin　cancer　emphasizes　the　need　for　quick　and　accurate　diagnosis　to　improve　patient　outcomes.　Although,　it　might　be　challenging　to　evaluate　the　possible　risk　of　a　skin　spot　merely　by　looking　at　it　and　feeling　it.　This　review　article　offers　a　thorough　overview　of　current　breakthroughs　in　machine　learning　(ML)　and　computer-aided　diagnostics　(CAD)　for　the　aim　of　analysis　and　classification　of　skin　cancer　lesions　over　the　past　6　years.　This　paper　carefully　reviews　the　whole　diagnostic　process:　data　preparation,　lesion　segmentation,　feature　extraction,　feature　selection,　and　final　classification.　Analyzed　are　many　publicly　accessible　datasets　and　creative　ideas　including　deep　learning　(DL)　and　ML　integrated　with　computer　vision,　together　with　their　impact　on　increasing　diagnosis　accuracy.　Given　the　variety　and　complexity　of　skin　lesions,　even　with　enormous　progress,　there　are　still　major　obstacles.　This　review　rigorously　assesses　current　methods,　notes　areas　of　great　challenge,　and　provides　recommendations　to　direct　the　next　research　targeted　at　improving　early　detection　strategies　and　CAD　systems.