xval-methods          package:MLInterfaces          R Documentation

_s_u_p_p_o_r_t _f_o_r _c_r_o_s_s-_v_a_l_i_d_a_t_o_r_y _m_a_c_h_i_n_e _l_e_a_r_n_i_n_g _w_i_t_h _e_x_p_r_S_e_t_s

_D_e_s_c_r_i_p_t_i_o_n:

     support for cross-validatory machine learning with exprSets

_M_e_t_h_o_d_s:


     _d_a_t_a = "_e_x_p_r_S_e_t", _c_l_a_s_s_L_a_b = "_c_h_a_r_a_c_t_e_r", _p_r_o_c = "_n_o_n_s_t_a_n_d_a_r_d_G_e_n_e_r_i_c", _x_v_a_l_M_e_t_h_o_d = "_c_h_a_r_a_c_t_e_r", _g_r_o_u_p = "_i_n_t_e_g_e_r" 'clas
          sLab' is the name of a component of the phenoData of the
          exprSet passed as 'data'. 'proc' is an actual MLInterfaces
          method (not the name of a method).  'xvalMethod' may have
          value "LOO" for leave-one-out or "LOG" for leave-group-out. 
          The latter makes use of the 'group' parameter.  samples
          sharing a value of 'group' are left out in one iteration of
          the cross-validation procedure, and predictions are made for
          them together on the basis of the fit from which they were
          excluded.

_E_x_a_m_p_l_e_s:

     library(golubEsets)
     data(golubMerge)
     smallG <- golubMerge[200:250,]
     lk1 <- xval(smallG, "ALL.AML", knnB, xvalMethod="LOO", group=as.integer(0))
     table(lk1,smallG$ALL.AML)
     lk2 <- xval(smallG, "ALL.AML", knnB, xvalMethod="LOG", group=as.integer(
      rep(1:8,each=9)))
     table(lk2,smallG$ALL.AML)

