search

accosmin / nano

C++ library [machine learning & numerical optimization] - superseeded by libnano
MIT License
1 stars 0 forks source link

Large scale benchmark of batch algorithms #69

Closed accosmin closed 8 years ago

accosmin commented 8 years ago

use only test functions with variable number of dimensions: 2,4,8,16,32,64 ..., 65K

accosmin commented 8 years ago

The maximum number of dimensions should be read from command line

accosmin commented 8 years ago

Implement a range of dimensions to consider: "--min-dims=1 --max-dims=10000". Thus can perform both small and large scale experiments. Add also the number of trials as a command line parameter.