tommy/libpostal
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[classification] Sparse version of logistic regression gradient which, given an array of the features/columns used in the input batch, only updates the gradient for that batch, even for the operations which otherwise would apply to the entire matrix (scaling by -1/m, regularization)

Browse Source
This commit is contained in:
Al
2016-01-09 01:33:33 -05:00
parent 5ca4bba1d5
commit 562cc06eaf
2 changed files with 76 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/logistic_regression.h
View File

@@ -29,5 +29,7 @@ may be called softmax regression.
bool logistic_regression_model_expectation(matrix_t *theta, sparse_matrix_t *x, matrix_t *p_y);
double logistic_regression_cost_function(matrix_t *theta, sparse_matrix_t *x, uint32_array *y, matrix_t *p_y, double lambda);
bool logistic_regression_gradient(matrix_t *theta, matrix_t *gradient, sparse_matrix_t *x, uint32_array *y, matrix_t *p_y, double lambda);
bool logistic_regression_gradient_sparse(matrix_t *theta, matrix_t *gradient, sparse_matrix_t *x, uint32_array *y, matrix_t *p_y,
uint32_array *x_cols, double lambda);
#endif