Predication accuracy of SGDLearner drops when the training data becomes large

huawei-noah / streamDM

Stream Data Mining Library for Spark Streaming

http://streamdm.noahlab.com.hk/

Apache License 2.0

492 stars 147 forks source link

Predication accuracy of SGDLearner drops when the training data becomes large #31

Closed roywjt closed 6 years ago

roywjt commented 8 years ago

When we ran the SGDLearner related algorithms(Perceptron, SGDLearner), the predication accuracy drops when the training data becomes large, for example millions.

Here we used local mode, the configuration is "Local[4]" and all the other parameters of the SGDLearner are set as default.