The　backward　heat　equation　is　a　typical　ill-posed　problem.　In　this　paper,　we　shall　apply　a　dual　least　squares　method　connecting　Shannon　wavelet　to　the　following　equation　{u(t)(x,y,t)　=　u(xx)(x,y,t)　+　u(yy)(x,y,t),　x　is　an　element　of　R,　y　is　an　element　of　R,　0　＜=　t　＜　1,　u(x,y,1)　=　phi(x,y),　x　is　an　element　of　R,　y　is　an　element　of　R.　Motivated　by　Reginska＇s　work,　we　shall　give　two　nonlinear　approximate　methods　to　regularize　the　approximate　solutions　for　high-dimensional　backward　heat　equation,　and　prove　that　our　methods　are　convergent.