Experimental Evaluation

	A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P
1
2	Dataset Used
3			Dataset: Agrawal 3.000.000 (total of 9 features => 2 out of 9 used)
4			agrawal_dataset_3M_abrupt.arff
5			expecting near 60.480.000			60,686,320
6	Monitoring Machine Learning Task After the Online Bagging
7					Total Execution Time
8	No. of Hoeffding Trees	Kafka Brokers	Partitions	Parallelism	Run 1		Run 2		Run 3		Run 4		Run 5		Mean Execution Time	Throughput
9	No. of Hoeffding Trees	Kafka Brokers	Partitions	Parallelism	Minutes	Sec	Minutes	Sec	Minutes	Sec	Minutes	Sec	Minutes	Sec
10	32	4	2	1											#DIV/0!	#DIV/0!
11	32	4	2	2	1h 22m	4920									4920	0.00164
12	32	4	8	4	23m 40s	1420	29m 2s	1742	32m 2s	1922	29m 4s	1744			1635.333333	0.0005451111111
13	32	4	8	6											#DIV/0!	#DIV/0!
14	32	4	8	8	17m 52s	1072	11m 8s	668	15m 6s	906	20m	1200	21m 41s	1301	984.5	0.0003281666667
15	32	4	8	12	13m 34s	814									814	0.0002713333333
16	32	4	8	16	12m 35s	755	9m 40s	580	8m 47s	527	10m 12s	612	8m 22s	502	591	0.000197
17	32	4	8	32	5m 21 s	321	4m 14s	254	5m 03s	303	4m 54s	294	5m 39s	339	304.25	0.0001014166667
18
19	Dataset Used
20			Dataset: Agrawal 30.000.000 (total of 9 features => 2 out of 9 used)
21
22
23	Monitoring Machine Learning Task After the Online Bagging
24			Total Execution Time
25	No. of Hoeffding Trees	Parallelism	Run 1	Run 2	Run 3	Run 4	Run 5	Mean Execution Time	Throughput
26	32	32
27
28
29
30
31
32
33	System Parameters (Default Values)
34	Parameter Name	Value	Comments
35	parallelism
36	number of Hoeffding Trees	32
37	Combination Function	3	top-k choice
38	Weighted Voting Parameter	3	No. of HT participating in Voting
39	age_of_maturity	10
40	drift_detection_method	2	VDDM
41	m_features	2	2 out of 9 = 36 possible distinct products
42	Max	9
43	max_examples_seen	200
44	delta	0.0000001
45	tie threshold	0.05
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100