[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

more possible model optimizations



1)
 zero-center (0-center) the normalized values in the model vectors, and then
 precalculate each feature vector element based on the offset, and then
 pass the sum to each support vector calculation.

 assumption: median is close to the mean.

2)
 cache individual values in the input feature vector, have 13000 values

 assumption: memory won't be too big




Why do you want this page removed?