An approach for transgender population information extraction and summarization from clinical trial text

Table 5 The performance comparison on the datasets (A to G) using Macro-averaged F₁-measure

Method	A	B	C	D	E	F	G
Logit Boost	0.637	0.674	0.681	0.639	0.667	0.636	0.628
Logistic	0.745	0.735	0.693	0.667	0.678	0.706	0.646
Bayes Net	0.680	0.662	0.652	0.624	0.665	0.665	0.655
Simple Logistic	0.761	0.668	0.697	0.684	0.644	0.685	0.658
LMT	0.772	0.668	0.643	0.686	0.625	0.686	0.665
Random Committee	0.728	0.738	0.696	0.695	0.688	0.750	0.673
Decision Table	0.637	0.609	0.590	0.599	0.605	0.617	0.675
Random Tree	0.674	0.667	0.661	0.646	0.652	0.668	0.718
Random Forest	0.774	0.739	0.760	0.698	0.733	0.747	0.765
Our approach	0.858	0.860	0.873	0.879	0.876	0.876	0.878

ISSN: 1472-6947