Distributed learning with bagging - like performance 3

10 Bagging forms a committee of classifiers by bootstrap aggregation of training sets from a pool of training data. A 11 simple alternative to bagging is to partition the data into disjoint subsets. Experiments with decision tree and neural 12 network classifiers on various datasets show that, given the same size partitions and bags, disjoint partitions result in 13 performance equivalent to, or better than, bootstrap aggregates (bags). Many applications (e.g., protein structure 14 prediction) involve use of datasets that are too large to handle in the memory of the typical computer. Hence, bagging 15 with samples the size of the data is impractical. Our results indicate that, in such applications, the simple approach of 16 creating a committee of n classifiers from disjoint partitions each of size 1=n (which will be memory resident during 17 learning) in a distributed way results in a classifier which has a bagging-like performance gain. The use of distributed 18 disjoint partitions in learning is significantly less complex and faster than bagging. 19 2002 Published by Elsevier Science B.V.