A comparative study on deep learning models for text classification of unstructured medical notes with various levels of class imbalance

Table 4 Number of samples in each class in training and test sets

Disease	Prevalence	Training Set		Test Set
Disease	Prevalence	Disease Presence Presence	Disease Absence	Disease Presence	Disease Absence
Hypertriglyceridemia	5%	50	878	17	292
Venous Insufficiency	7%	62	865	21	289
Asthma	13%	123	805	41	268
Gout	13%	120	808	40	269
OSA	14%	129	799	43	266
PVD	15%	135	793	45	264
Gallstones	15%	141	787	47	262
OA	18%	168	760	56	253
GERD	20%	184	743	62	248
Depression	20%	187	741	62	247
Obesity	40%	374	554	125	184
CHF	43%	402	526	134	175
Hypercholesterolemia	47%	432	496	144	165
CAD	55%	512	416	170	139
Diabetes	66%	616	312	205	104
Hypertension	73%	677	250	226	84

ISSN: 1471-2288