Used the algorithm is Stochastic Gradient Descent (SGD). Used the library is pyspark.mllib. Set the step_size=10 and mini_batch_fraction=1, draw the graph for the numIterations. The AUC is the highest ...
serialized_pb=_b('\n\x0e\x64ist_sgd.proto\x12\x08\x64ist_sgd\"`\n\tSubTensor\x12\x12\n\ntensor_len\x18\x01 \x01(\x05\x12\x14\n\x0ctensor_chunk\x18\x02 \x01(\x05\x12 ...