Let
class="mathmlsrc">class="formulatext stixSupport mathImg" data-mathURL="/science?_ob=MathURL&_method=retrieve&_eid=1-s2.0-S0022039616301760&_mathId=si1.gif&_user=111111111&_pii=S0022039616301760&_rdoc=1&_issn=00220396&md5=5c0550a174b7c2b603ee8dfed4fab9af" title="Click to view the MathML source">G=(V,E)class="mathContainer hidden">class="mathCode"> be a locally finite graph,
class="mathmlsrc">class="formulatext stixSupport mathImg" data-mathURL="/science?_ob=MathURL&_method=retrieve&_eid=1-s2.0-S0022039616301760&_mathId=si2.gif&_user=111111111&_pii=S0022039616301760&_rdoc=1&_issn=00220396&md5=b055c95fa158e14af49b3ea4228aad65" title="Click to view the MathML source">Ω⊂Vclass="mathContainer hidden">class="mathCode"> be a bounded domain, Δ be
the usual graph Laplacian, and
class="mathmlsrc">class="formulatext stixSupport mathImg" data-mathURL="/science?_ob=MathURL&_method=retrieve&_eid=1-s2.0-S0022039616301760&_mathId=si3.gif&_user=111111111&_pii=S0022039616301760&_rdoc=1&_issn=00220396&md5=1cc12c5e63b17b058342a7810b75845e" title="Click to view the MathML source">λ1(Ω)class="mathContainer hidden">class="mathCode"> be
the first eigenvalue of −Δ with respect to Dirichlet boundary condition. Using
the mountain pass
theorem due to Ambrosetti–Rabinowitz, we prove that if
class="mathmlsrc">class="formulatext stixSupport mathImg" data-mathURL="/science?_ob=MathURL&_method=retrieve&_eid=1-s2.0-S0022039616301760&_mathId=si220.gif&_user=111111111&_pii=S0022039616301760&_rdoc=1&_issn=00220396&md5=9e803a4c36d0ead96660bdac88fb0b85" title="Click to view the MathML source">α<λ1(Ω)class="mathContainer hidden">class="mathCode">,
then for any
class="mathmlsrc">class="formulatext stixSupport mathImg" data-mathURL="/science?_ob=MathURL&_method=retrieve&_eid=1-s2.0-S0022039616301760&_mathId=si5.gif&_user=111111111&_pii=S0022039616301760&_rdoc=1&_issn=00220396&md5=65e8fdd158b842879d6c79a28527a8a9" title="Click to view the MathML source">p>2class="mathContainer hidden">class="mathCode">,
there exists a positive solution to
class="formula" id="fm0010">
where
class="mathmlsrc">class="formulatext stixSupport mathImg" data-mathURL="/science?_ob=MathURL&_method=retrieve&_eid=1-s2.0-S0022039616301760&_mathId=si231.gif&_user=111111111&_pii=S0022039616301760&_rdoc=1&_issn=00220396&md5=dcdc6edb867fd635a3cfd08febc662d7" title="Click to view the MathML source">Ω∘class="mathContainer hidden">class="mathCode"> and ∂Ω denote
the interior and
the boundary of Ω respectively. Also we consider similar problems involving
the p-Laplacian and poly-Laplacian by
the same method. Such problems can be viewed as discrete versions of
the Yamabe type equations on Euclidean space or compact Riemannian manifolds.