Table 4.

CAD: Comparison of F1 measure for CAD models between 2008, 2014 and combined 2008/2014 broken down by SVM kernel, weighting scheme, and feature space.

Kernel	Weight	Feature Space	2008	2014	Combined 2008/2014	% Change 2008, 2014	% Change 2008, Comb.	% Change 2014, Comb.
RBF	tf-idf	Min	0.841††	0.896	0.878	6.54**	4.37**	−2.03
RBF	tf-idf	Min, Lex & CUI	0.827††	0.891	0.872	7.73**	5.47**	−2.10
RBF	tf-idf	Min & Lex	0.829††	0.883	0.871†	6.62**	5.12**	−1.41
Linear	tf-idf	Min & Lex	0.805††	0.882	0.843††	9.62**	4.77**	−4.42**
Linear	tf-idf	Lex	0.815††	0.881	0.844††	8.09**	3.47**	−4.28**
RBF	tf-idf	Lex	0.748††	0.875†	0.876	17.01**	17.09**	0.07
Linear	tf-idf	Min, Lex & CUI	0.793††	0.875	0.846††	10.34**	6.72**	−3.28*
Linear	tf-idf	Min	0.809††	0.872††	0.841††	7.87**	3.96**	−3.63**
Linear	tf-idf	Min & CUI	0.788††	0.868††	0.861††	10.15**	9.30**	−0.77
Linear	tf-idf	Lex & CUI	0.780††	0.863††	0.851††	10.60**	9.06**	−1.39
RBF	tf-idf	Min & CUI	0.820††	0.863††	0.868†	5.16**	5.77**	0.58
RBF	Count	Min	0.727††	0.861††	0.830††	18.40**	14.10**	−3.63
RBF	Count	Lex	0.792††	0.858††	0.833††	8.37**	5.14**	−2.97*
RBF	Count	Lex & CUI	0.765††	0.857††	0.837††	12.00**	9.41**	−2.31
RBF	tf-idf	Lex & CUI	0.782††	0.855††	0.881	9.29**	12.68**	3.10
Linear	Count	Lex & CUI	0.737††	0.851††	0.791††	15.51**	7.33**	−7.09**
RBF	Count	Min & CUI	0.696††	0.851††	0.830††	22.23**	19.16**	−2.51
RBF	Count	Min & Lex	0.670††	0.844††	0.843††	25.87**	25.73**	−0.11
Linear	Count	Lex	0.698††	0.841††	0.814††	20.47**	16.66**	−3.16
Linear	Count	Min, Lex & CUI	0.698††	0.840††	0.767††	20.32**	9.86**	−8.69**
Linear	Count	Min & Lex	0.674††	0.835††	0.799††	23.89**	18.51**	−4.34*
RBF	Count	Min, Lex & CUI	0.714††	0.832††	0.842††	16.57**	17.89**	1.13
Linear	Count	Min	0.682††	0.825††	0.795††	21.05**	16.62**	−3.66
Linear	Count	Min & CUI	0.669††	0.824††	0.797††	23.30**	19.20**	−3.32
RBF	tf-idf	CUI	0.608††	0.689††	0.732††	13.29**	20.43**	6.30*
Linear	tf-idf	CUI	0.610††	0.674††	0.654††	10.65**	7.31*	−3.01
RBF	Count	CUI	0.631††	0.616††	0.616††	−2.41	−2.46	−0.04
Linear	Count	CUI	0.542††	0.538††	0.548††	−0.65	1.23	1.89

^*percent change significant at alpha = 0.05

^**percent change significant at alpha = 0.1

^†statistically significantly different from highest performing model (emphasized above) at alpha = 0.05

^††statistically significantly different from highest performing model (emphasized above) at alpha =0.1

Note: For visual reference, models which are not statistically significantly different from the emphasized model at alpha = 0.1 are shaded in dark grey. Models which are not statistically significantly different from the emphasized model at alpha = 0.05 are shaded in light grey