Internals of svm_predict¶
This document describes the structure and execution of svm prediction, implemented as svmPredict
function.
Svm_predict is a function to predict sample’s classification by trained weight vector and sample feature vector. This function provide a stream in, stream out module to easily get prediction class of sample.
The structure of svmPredict
is described as below. The primitive have two function which are getWeight
and getPredictions
.
The input of getWeight
is stream and this part stores weight vector in ram.
getPredictions
has a dataflow region including two functions. one is dot_multiply, the other is tag type transform.
The hardware resource utilization of svm_prediction is shown in the table below (work as 300MHz).
LUT | FF | BRAM | URAM | DSP |
27114 | 37335 | 26 | 0 | 133 |