Multiple Instance Learning Datasets

This page aims at providing to the evolutionary learning/machine learning researchers a set of MIL bechmarks to analyze the behavior of the learning methods. Both data and partitions of data set for MIL different applications are available.

Dataset Attributes
Bags
Instances Average Bag size
Positive Negative Total
Drug Activity Prediction
Musk1
166
47
45
92
476
5.17
Musk2
166
39
63
102
6598
64.69
Mutagenesis_Atoms
10
125
63
188
1618
8.61
Mutagenesis_Bonds
16
125
63
188
3995
21.25
Mutagenesis_Chains
24
125
63
188
5349
28.45
Content-based image retrieval and classification
Tiger
230
100
100
200
1391
6.96
Elephant
230
100
100
200
1220
6.10
Fox
320
100
100
200
1320
6.60