Multiclass Classification by Various Machine Learning Algorithms and Interpretation of the Risk Factors of Pedestrian Accidents Using Explainable AI

Table 3

Classification prediction results of pedestrian accident data.

Method	Severity level	HPO	Metrics				HPO	Metrics
			Accuracy	Recall	Precision	F1-score		Accuracy	Recall	Precision	F1-score
LR	0	Default	0.677	0.433	0.665	0.525	GS	0.677	0.433	0.666	0.525
	1			0.364	0.288	0.322			0.364	0.288	0.322
	2			0.008	0.014	0.011			0.008	0.014	0.010
NB	0		0.517	0.62	0.57	0.60		0.517	0.625	0.574	0.598
	1			0.15	0.27	0.20			0.154	0.273	0.197
	2			0.91	0.18	0.31			0.915	0.185	0.307
XGB	0		0.700	0.50	0.67	0.57		0.705	0.507	0.677	0.580
	1			0.32	0.27	0.39			0.316	0.270	0.291
	2			0.02	0.02	0.02			0.015	0.016	0.015
LGBM	0		0.708	0.50	0.69	0.58		0.705	0.509	0.677	0.581
	1			0.32	0.27	0.28			0.315	0.270	0.291
	2			0.01	0.02	0.02			0.016	0.017	0.017
CB	0		0.700	0.50	0.67	0.57		0.704	0.487	0.686	0.570
	1			0.32	0.27	0.29			0.329	0.276	0.300
	2			0.01	0.02	0.02			0.016	0.017	0.017
LR	0	RS	0.687	0.460	0.670	0.545	BO	0.688	0.460	0.670	0.546
	1			0.346	0.284	0.312			0.346	0.284	0.312
	2			0.019	0.024	0.021			0.018	0.022	0.020
NB	0		0.516	0.625	0.573	0.598		0.577	0.574	0.598	0.586
	1			0.332	0.273	0.197			0.217	0.282	0.246
	2			0.011	0.185	0.307			0.593	0.162	0.254
XGB	0		0.702	0.482	0.689	0.567		0.704	0.501	0.678	0.576
	1			0.332	0.275	0.301			0.321	0.272	0.294
	2			0.011	0.013	0.012			0.010	0.011	0.010
LGBM	0		0.707	0.500	0.678	0.576		0.708	0.502	0.687	0.580
	1			0.321	0.273	0.295			0.319	0.271	0.293
	2			0.016	0.017	0.017			0.015	0.016	0.015
CB	0		0.707	0.499	0.688	0.579		0.706	0.514	0.677	0.584
	1			0.321	0.270	0.293			0.311	0.267	0.288
	2			0.013	0.015	0.014			0.013	0.015	0.014

Four metrics are used to evaluate the model’s performance.