Comparison of decision tree with common machine learning models for prediction of biguanide and sulfonylurea poisoning in the United States: an analysis of the National Poison Data System

Mehrpour, Omid; Saeedi, Farhad; Nakhaee, Samaneh; Tavakkoli Khomeini, Farbod; Hadianfar, Ali; Amirabadizadeh, Alireza; Hoyte, Christopher

doi:10.1186/s12911-022-02095-y

BMC Medical Informatics and Decision Making

Table 3 Characteristics of ML models used in comparative analysis

From: Comparison of decision tree with common machine learning models for prediction of biguanide and sulfonylurea poisoning in the United States: an analysis of the National Poison Data System

Labels	ML models	Biguanides	Sulfonylurea	Average	Weighted_average
Specificity	Adaboosting	0.907973	0.954436	0.931205	0.929368
	DT	0.898138	0.963429	0.930784	0.928202
	SGD	0.790657	0.960731	0.875694	0.868969
	SVM_linear	0.896382	0.962230	0.929306	0.926702
	MLP	0.913593	0.952338	0.932966	0.931434
	Gradient boosting	0.901300	0.969125	0.935212	0.932530
	Light gradient boosting	0.909730	0.954736	0.932233	0.930453
	Voting-ensemble	0.895328	0.972422	0.933875	0.930827
	Bagging ensemble	0.892870	0.973022	0.932946	0.929776
	Stacking ensemble	0.893572	0.973321	0.933447	0.930293
Precision	Adaboosting	0.923970	0.944465	0.934217	0.933407
	DT	0.917237	0.954461	0.935849	0.934377
	SGD	0.843199	0.945004	0.894102	0.890076
	SVM_linear	0.915835	0.952950	0.934392	0.932925
	MLP	0.928133	0.942391	0.935262	0.934698
	Gradient boosting	0.920034	0.961409	0.940721	0.939085
	Light gradient boosting	0.925334	0.944911	0.935122	0.934348
	Voting-ensemble	0.915867	0.965165	0.940516	0.938566
	Bagging ensemble	0.914109	0.965805	0.939957	0.937913
	Stacking ensemble	0.914648	0.966198	0.940423	0.938385
Recall	Adaboosting	0.954436	0.907973	0.931205	0.933042
	DT	0.963429	0.898138	0.930784	0.933366
	SGD	0.960731	0.790657	0.875694	0.882420
	SVM_linear	0.962230	0.896382	0.929306	0.931910
	MLP	0.952338	0.913593	0.932966	0.934498
	Gradient boosting	0.969125	0.901300	0.935212	0.937894
	Light gradient boosting	0.954736	0.909730	0.932233	0.934013
	Voting-ensemble	0.972422	0.895328	0.933875	0.936924
	Bagging ensemble	0.973022	0.892870	0.932946	0.936115
	Stacking ensemble	0.973321	0.893572	0.933447	0.936600
F1_score	Adaboosting	0.938956	0.925860	0.932408	0.932926
	DT	0.939766	0.925443	0.932605	0.933171
	SGD	0.898136	0.860968	0.879552	0.881022
	SVM_linear	0.938459	0.923801	0.931130	0.931710
	MLP	0.940080	0.927769	0.933924	0.934411
	Gradient boosting	0.943942	0.930384	0.937163	0.937699
	Light gradient boosting	0.939805	0.926986	0.933396	0.933903
	Voting-ensemble	0.943297	0.928936	0.936117	0.936685
	Bagging ensemble	0.942646	0.927907	0.935276	0.935859
	Stacking ensemble	0.943073	0.928467	0.935770	0.936348
Accuracy	Adaboosting	–	–	0.933042	0.933042
	DT	–	–	0.933366	0.933366
	SGD	–	–	0.882420	0.882420
	SVM_linear	–	–	0.931910	0.931910
	MLP	–	–	0.934498	0.934498
	Gradient boosting	–	–	0.937894	0.937894
	Light gradient boosting	–	–	0.934013	0.934013
	Voting-ensemble	–	–	0.936924	0.936924
	Bagging ensemble	–	–	0.936115	0.936115
	Stacking ensemble	–	–	0.936600	0.936600

DT Decision tree, MLP Multi layers perceptron, SGD Stochastic gradient descent, Adaboosting classiefier, SVM_linear: linear support vector machine

Back to article page

ISSN: 1472-6947

Contact us

General enquiries: journalsubmissions@springernature.com