The impact of chemotherapy and survival prediction by machine learning in early Elderly Triple Negative Breast Cancer (eTNBC): a population based study from the SEER database

Table 7 Confusion matrix of nine algorithms for 5-year BCSS and 5-year OS

Algorithms		Predictions		Algorithms		Predictions
Algorithms		Dead	Alive	Algorithms		Dead	Alive
5-year BCSS				5-year OS
K-nearest neighbor	Dead	3	47	K-nearest neighbor	Dead	15	56
K-nearest neighbor	Alive	7	350	K-nearest neighbor	Alive	17	337
Catboost	Dead	8	42	Catboost	Dead	15	56
Catboost	Alive	9	348	Catboost	Alive	8	346
Decision tree	Dead	13	37	Decision tree	Dead	21	50
Decision tree	Alive	18	339	Decision tree	Alive	21	333
Random forest	Dead	7	43	Random forest	Dead	18	53
Random forest	Alive	10	347	Random forest	Alive	16	338
Gradient booster	Dead	5	45	Gradient booster	Dead	12	59
Gradient booster	Alive	3	354	Gradient booster	Alive	5	349
LightGBM	Dead	5	45	LightGBM	Dead	14	57
LightGBM	Alive	3	354	LightGBM	Alive	6	348
Neural network model	Dead	0	50	Neural network model	Dead	24	47
Neural network model	Alive	0	357	Neural network model	Alive	18	336
Support vector machine	Dead	5	45	Support vector machine	Dead	11	60
Support vector machine	Alive	3	354	Support vector machine	Alive	2	352
XGBoost	Dead	8	42	XGBoost	Dead	18	53
XGBoost	Alive	7	350	XGBoost	Alive	4	350

ISSN: 1471-2318