A Bayesian Network Approach to Causation Analysis of Road Accidents Using Netica

Table 2

Variables used for the Construction of Bayesian Network.

Variable class	Variable name	Discretization value	Value description	Frequency	Percentage
Driver	Apparent error	1	Fail to stand	199	21.47%
		2	Change lanes to endanger	141	15.21%
		3	Incorrect turn	24	2.59%
		4	Reverse without due care	61	6.58%
		5	Follow too closely	119	12.84%
		6	Overtake without due care	35	3.78%
		7	Disobey traffic lights	115	12.41%
		8	Disobey stop sign	18	1.94%
		9	Disobey give way sign	27	2.91%
		10	DUI	9	0.97%
		11	Fail to give way	179	19.31%
Road	Road geometry	1	Cross road	454	48.98%
		2	Y junction	11	1.19%
		3	T junction	196	21.14%
		4	Multiple	13	1.40%
		5	Divided road	132	14.24%
		6	Not divided	121	13.05%
	Road moisture condition	1	Wet	100	10.79%
		2	Dry	827	89.21%
	Traffic control	1	Traffic signals	477	51.46%
		2	Stop sign	27	2.91%
		3	Give way sign	56	6.04%
		4	No control	367	39.59%
Environment	Weather condition	1	Raining	65	7.01%
		2	Not raining	862	92.99%
	Light condition	1	Daylight	677	73.03%
		2	Night	250	26.97%
Vehicle	Vehicle type	1	Heavy	63	6.80%
		2	Medium	379	40.88%
		3	Light	485	52.32%
	Vehicle movement	1	Right turn	267	28.80%
		2	Left turn	41	4.42%
		3	U turn	79	8.52%
		4	Swerving	137	14.78%
		5	Reversing	52	5.61%
		6	Straight ahead	275	29.67%
		7	Entering private driveway	10	1.08%
		8	Leaving private driveway	31	3.34%
		9	Overtaking on right	25	2.70%
		10	Overtaking on left	10	1.08%
Road crash	Crash type	1	Rear end	155	16.72%
		2	Hit fixed object	4	0.43%
		3	Side swipe	255	27.51%
		4	Right angle	258	27.83%
		5	Head on	3	0.32%
		6	Hit pedestrian	22	2.37%
		7	Right turn	216	23.30%
		8	Hit parked vehicle	14	1.51%
	Crash severity	1	PDO (property damage only)	689	74.33%
		2	Injury	238	25.67%
	Total units (involved in a road crash)	1	Two units	860	92.77%
		2	Three units	54	5.83%
		3	Four units	9	0.97%
		4	Five units	4	0.43%
	Total casualties (fatalities and treated injuries)	1	None	689	74.33%
		2	One casualty	213	22.98%
		3	Two casualties	21	2.27%
		4	Three casualties	4	0.43%
	Total serious injuries	1	None	904	97.52%
		2	One serious injury	23	2.48%
	Total estimated damage (A$)	1	[0, 5000)	423	45.63%
		2	[5000, 10000)	342	36.89%
		3	[10000, +∞)	162	17.48%