Cordey, J. G.Thomsen, K.Chudnovskiy, A.Kardaun, OjwfTakizuka, T.Snipes, J. A.Greenwald, M.Sugiyama, L.Ryter, F.Kus, A.Stober, J.DeBoo, J. C.Petty, C. C.Bracco, G.Romanelli, M.Cui, Z.Liu, Y.McDonald, D. C.Meakins, A.Miura, Y.Shinohara, K.Tsuzuki, K.Kamada, Y.Urano, H.Valovic, M.Akers, R.Brickley, C.Sykes, A.Walsh, M. J.Kaye, S. M.Bush, C.Hogewei, D.Martin, Y.Cote, A.Pacher, G.Ongena, J.Imbeaux, F.Hoang, G. T.Lebedev, S.Leonov, V.2008-04-162008-04-162008-04-16200510.1088/0029-5515/45/9/007https://infoscience.epfl.ch/handle/20.500.14299/22227WOS:0002323501000077445The condition of the latest version of the ELMy H-mode database has been re-examined. It is shown that there is bias in the ordinary least squares regression for some of the variables. To address these shortcomings three different techniques are employed: (a) principal component regression, (b) an error in variables technique and (c) the selection of a better conditioned dataset with fewer variables. Scalings in terms of the dimensionless physics variables, as well as the standard set of engineering variables, are also derived. The new scalings give a very similar performance for existing scalings for ITER at the standard beta(n) of 1.6, but a much improved performance at higher beta n.ITERtext::journal::journal article::research article