mirror of
https://github.com/Relintai/pmlpp.git
synced 2024-12-21 14:56:47 +01:00
Removed new things from MLPPUtilitiesOld.
This commit is contained in:
parent
80f85ee0af
commit
d6037730b5
@ -6,10 +6,6 @@
|
||||
|
||||
#include "utilities_old.h"
|
||||
|
||||
#include "core/log/logger.h"
|
||||
#include "core/math/math_funcs.h"
|
||||
#include "core/math/random_pcg.h"
|
||||
|
||||
#include <fstream>
|
||||
#include <iostream>
|
||||
#include <random>
|
||||
@ -109,176 +105,6 @@ std::vector<real_t> MLPPUtilitiesOld::biasInitialization(int n) {
|
||||
return bias;
|
||||
}
|
||||
|
||||
void MLPPUtilitiesOld::weight_initializationv(Ref<MLPPVector> weights, WeightDistributionType type) {
|
||||
ERR_FAIL_COND(!weights.is_valid());
|
||||
|
||||
int n = weights->size();
|
||||
real_t *weights_ptr = weights->ptrw();
|
||||
|
||||
RandomPCG rnd;
|
||||
rnd.randomize();
|
||||
|
||||
std::random_device rd;
|
||||
std::default_random_engine generator(rd());
|
||||
|
||||
switch (type) {
|
||||
case WEIGHT_DISTRIBUTION_TYPE_DEFAULT: {
|
||||
std::uniform_real_distribution<real_t> distribution(0, 1);
|
||||
|
||||
for (int i = 0; i < n; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_XAVIER_NORMAL: {
|
||||
std::normal_distribution<real_t> distribution(0, Math::sqrt(2.0 / (n + 1.0)));
|
||||
|
||||
for (int i = 0; i < n; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_XAVIER_UNIFORM: {
|
||||
std::uniform_real_distribution<real_t> distribution(-Math::sqrt(6.0 / (n + 1.0)), Math::sqrt(6.0 / (n + 1.0)));
|
||||
|
||||
for (int i = 0; i < n; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_HE_NORMAL: {
|
||||
std::normal_distribution<real_t> distribution(0, Math::sqrt(2.0 / n));
|
||||
|
||||
for (int i = 0; i < n; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_HE_UNIFORM: {
|
||||
std::uniform_real_distribution<real_t> distribution(-Math::sqrt(6.0 / n), Math::sqrt(6.0 / n));
|
||||
|
||||
for (int i = 0; i < n; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_LE_CUN_NORMAL: {
|
||||
std::normal_distribution<real_t> distribution(0, Math::sqrt(1.0 / n));
|
||||
|
||||
for (int i = 0; i < n; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_LE_CUN_UNIFORM: {
|
||||
std::uniform_real_distribution<real_t> distribution(-Math::sqrt(3.0 / n), Math::sqrt(3.0 / n));
|
||||
|
||||
for (int i = 0; i < n; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_UNIFORM: {
|
||||
std::uniform_real_distribution<real_t> distribution(-1.0 / Math::sqrt(static_cast<real_t>(n)), 1.0 / Math::sqrt(static_cast<real_t>(n)));
|
||||
|
||||
for (int i = 0; i < n; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
default:
|
||||
break;
|
||||
}
|
||||
}
|
||||
void MLPPUtilitiesOld::weight_initializationm(Ref<MLPPMatrix> weights, WeightDistributionType type) {
|
||||
ERR_FAIL_COND(!weights.is_valid());
|
||||
|
||||
int n = weights->size().x;
|
||||
int m = weights->size().y;
|
||||
int data_size = weights->data_size();
|
||||
real_t *weights_ptr = weights->ptrw();
|
||||
|
||||
RandomPCG rnd;
|
||||
rnd.randomize();
|
||||
|
||||
std::random_device rd;
|
||||
std::default_random_engine generator(rd());
|
||||
|
||||
switch (type) {
|
||||
case WEIGHT_DISTRIBUTION_TYPE_DEFAULT: {
|
||||
std::uniform_real_distribution<real_t> distribution(0, 1);
|
||||
|
||||
for (int i = 0; i < data_size; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_XAVIER_NORMAL: {
|
||||
std::normal_distribution<real_t> distribution(0, sqrt(2 / (n + m)));
|
||||
|
||||
for (int i = 0; i < data_size; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_XAVIER_UNIFORM: {
|
||||
std::uniform_real_distribution<real_t> distribution(-sqrt(6 / (n + m)), sqrt(6 / (n + m)));
|
||||
|
||||
for (int i = 0; i < data_size; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_HE_NORMAL: {
|
||||
std::normal_distribution<real_t> distribution(0, sqrt(2 / n));
|
||||
|
||||
for (int i = 0; i < data_size; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_HE_UNIFORM: {
|
||||
std::uniform_real_distribution<real_t> distribution(-sqrt(6 / n), sqrt(6 / n));
|
||||
|
||||
for (int i = 0; i < data_size; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_LE_CUN_NORMAL: {
|
||||
std::normal_distribution<real_t> distribution(0, sqrt(1 / n));
|
||||
|
||||
for (int i = 0; i < data_size; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_LE_CUN_UNIFORM: {
|
||||
std::uniform_real_distribution<real_t> distribution(-sqrt(3 / n), sqrt(3 / n));
|
||||
|
||||
for (int i = 0; i < data_size; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
case WEIGHT_DISTRIBUTION_TYPE_UNIFORM: {
|
||||
std::uniform_real_distribution<real_t> distribution(-1 / sqrt(n), 1 / sqrt(n));
|
||||
|
||||
for (int i = 0; i < data_size; ++i) {
|
||||
weights_ptr[i] = distribution(generator);
|
||||
}
|
||||
} break;
|
||||
default:
|
||||
break;
|
||||
}
|
||||
}
|
||||
real_t MLPPUtilitiesOld::bias_initializationr() {
|
||||
std::random_device rd;
|
||||
std::default_random_engine generator(rd());
|
||||
std::uniform_real_distribution<real_t> distribution(0, 1);
|
||||
|
||||
return distribution(generator);
|
||||
}
|
||||
void MLPPUtilitiesOld::bias_initializationv(Ref<MLPPVector> z) {
|
||||
ERR_FAIL_COND(!z.is_valid());
|
||||
|
||||
std::vector<real_t> bias;
|
||||
std::random_device rd;
|
||||
std::default_random_engine generator(rd());
|
||||
std::uniform_real_distribution<real_t> distribution(0, 1);
|
||||
|
||||
int n = z->size();
|
||||
|
||||
for (int i = 0; i < n; i++) {
|
||||
bias.push_back(distribution(generator));
|
||||
}
|
||||
}
|
||||
|
||||
real_t MLPPUtilitiesOld::performance(std::vector<real_t> y_hat, std::vector<real_t> outputSet) {
|
||||
real_t correct = 0;
|
||||
for (uint32_t i = 0; i < y_hat.size(); i++) {
|
||||
@ -305,50 +131,6 @@ real_t MLPPUtilitiesOld::performance(std::vector<std::vector<real_t>> y_hat, std
|
||||
return correct / y_hat.size();
|
||||
}
|
||||
|
||||
real_t MLPPUtilitiesOld::performance_vec(const Ref<MLPPVector> &y_hat, const Ref<MLPPVector> &output_set) {
|
||||
ERR_FAIL_COND_V(!y_hat.is_valid(), 0);
|
||||
ERR_FAIL_COND_V(!output_set.is_valid(), 0);
|
||||
|
||||
real_t correct = 0;
|
||||
for (int i = 0; i < y_hat->size(); i++) {
|
||||
if (Math::is_equal_approx(y_hat->get_element(i), output_set->get_element(i))) {
|
||||
correct++;
|
||||
}
|
||||
}
|
||||
return correct / y_hat->size();
|
||||
}
|
||||
real_t MLPPUtilitiesOld::performance_mat(const Ref<MLPPMatrix> &y_hat, const Ref<MLPPMatrix> &y) {
|
||||
ERR_FAIL_COND_V(!y_hat.is_valid(), 0);
|
||||
ERR_FAIL_COND_V(!y.is_valid(), 0);
|
||||
|
||||
real_t correct = 0;
|
||||
for (int i = 0; i < y_hat->size().y; i++) {
|
||||
int sub_correct = 0;
|
||||
|
||||
for (int j = 0; j < y_hat->size().x; j++) {
|
||||
if (Math::round(y_hat->get_element(i, j)) == y->get_element(i, j)) {
|
||||
sub_correct++;
|
||||
}
|
||||
|
||||
if (sub_correct == y_hat->size().x) {
|
||||
correct++;
|
||||
}
|
||||
}
|
||||
}
|
||||
return correct / y_hat->size().y;
|
||||
}
|
||||
real_t MLPPUtilitiesOld::performance_pool_int_array_vec(PoolIntArray y_hat, const Ref<MLPPVector> &output_set) {
|
||||
ERR_FAIL_COND_V(!output_set.is_valid(), 0);
|
||||
|
||||
real_t correct = 0;
|
||||
for (int i = 0; i < y_hat.size(); i++) {
|
||||
if (y_hat[i] == Math::round(output_set->get_element(i))) {
|
||||
correct++;
|
||||
}
|
||||
}
|
||||
return correct / y_hat.size();
|
||||
}
|
||||
|
||||
void MLPPUtilitiesOld::saveParameters(std::string fileName, std::vector<real_t> weights, real_t bias, bool app, int layer) {
|
||||
std::string layer_info = "";
|
||||
std::ofstream saveFile;
|
||||
@ -478,36 +260,6 @@ void MLPPUtilitiesOld::UI(std::vector<real_t> weights, std::vector<real_t> initi
|
||||
std::cout << bias << std::endl;
|
||||
}
|
||||
|
||||
void MLPPUtilitiesOld::print_ui_vb(Ref<MLPPVector> weights, real_t bias) {
|
||||
String str = "Values of the weight(s):\n";
|
||||
str += weights->to_string();
|
||||
str += "\nValue of the bias:\n";
|
||||
str += String::num(bias);
|
||||
|
||||
PLOG_MSG(str);
|
||||
}
|
||||
void MLPPUtilitiesOld::print_ui_vib(Ref<MLPPVector> weights, Ref<MLPPVector> initial, real_t bias) {
|
||||
String str = "Values of the weight(s):\n";
|
||||
str += weights->to_string();
|
||||
|
||||
str += "\nValues of the initial(s):\n";
|
||||
str += initial->to_string();
|
||||
|
||||
str += "\nValue of the bias:\n";
|
||||
str += String::num(bias);
|
||||
|
||||
PLOG_MSG(str);
|
||||
}
|
||||
void MLPPUtilitiesOld::print_ui_mb(Ref<MLPPMatrix> weights, Ref<MLPPVector> bias) {
|
||||
String str = "Values of the weight(s):\n";
|
||||
str += weights->to_string();
|
||||
|
||||
str += "\nValue of the biased:\n";
|
||||
str += bias->to_string();
|
||||
|
||||
PLOG_MSG(str);
|
||||
}
|
||||
|
||||
void MLPPUtilitiesOld::CostInfo(int epoch, real_t cost_prev, real_t Cost) {
|
||||
std::cout << "-----------------------------------" << std::endl;
|
||||
std::cout << "This is epoch: " << epoch << std::endl;
|
||||
@ -516,14 +268,6 @@ void MLPPUtilitiesOld::CostInfo(int epoch, real_t cost_prev, real_t Cost) {
|
||||
std::cout << Cost << std::endl;
|
||||
}
|
||||
|
||||
void MLPPUtilitiesOld::cost_info(int epoch, real_t cost_prev, real_t cost) {
|
||||
String str = "This is epoch: " + itos(epoch) + ",";
|
||||
str += "The cost function has been minimized by " + String::num(cost_prev - cost);
|
||||
str += ", Current Cost:" + String::num(cost);
|
||||
|
||||
PLOG_MSG(str);
|
||||
}
|
||||
|
||||
std::vector<std::vector<std::vector<real_t>>> MLPPUtilitiesOld::createMiniBatches(std::vector<std::vector<real_t>> inputSet, int n_mini_batch) {
|
||||
int n = inputSet.size();
|
||||
|
||||
@ -598,189 +342,6 @@ std::tuple<std::vector<std::vector<std::vector<real_t>>>, std::vector<std::vecto
|
||||
return { inputMiniBatches, outputMiniBatches };
|
||||
}
|
||||
|
||||
Vector<Ref<MLPPMatrix>> MLPPUtilitiesOld::create_mini_batchesm(const Ref<MLPPMatrix> &input_set, int n_mini_batch) {
|
||||
Size2i size = input_set->size();
|
||||
|
||||
int n = size.y;
|
||||
int mini_batch_element_count = n / n_mini_batch;
|
||||
|
||||
Ref<MLPPVector> row_tmp;
|
||||
row_tmp.instance();
|
||||
row_tmp->resize(size.x);
|
||||
|
||||
Vector<Ref<MLPPMatrix>> input_mini_batches;
|
||||
|
||||
// Creating the mini-batches
|
||||
for (int i = 0; i < n_mini_batch; i++) {
|
||||
int mini_batch_start_offset = n_mini_batch * i;
|
||||
Ref<MLPPMatrix> current_input_set;
|
||||
current_input_set.instance();
|
||||
current_input_set->resize(Size2i(size.x, mini_batch_element_count));
|
||||
|
||||
for (int j = 0; j < mini_batch_element_count; j++) {
|
||||
input_set->get_row_into_mlpp_vector(mini_batch_start_offset + j, row_tmp);
|
||||
current_input_set->set_row_mlpp_vector(j, row_tmp);
|
||||
}
|
||||
|
||||
input_mini_batches.push_back(current_input_set);
|
||||
}
|
||||
|
||||
/* Don't think this can ever happen, todo double check
|
||||
if (real_t(n) / real_t(n_mini_batch) - int(n / n_mini_batch) != 0) {
|
||||
for (int i = 0; i < n - n / n_mini_batch * n_mini_batch; i++) {
|
||||
inputMiniBatches[n_mini_batch - 1].push_back(inputSet[n_mini_batch * n_mini_batch + i]);
|
||||
}
|
||||
}
|
||||
*/
|
||||
|
||||
return input_mini_batches;
|
||||
}
|
||||
MLPPUtilitiesOld::CreateMiniBatchMVBatch MLPPUtilitiesOld::create_mini_batchesmv(const Ref<MLPPMatrix> &input_set, const Ref<MLPPVector> &output_set, int n_mini_batch) {
|
||||
Size2i size = input_set->size();
|
||||
|
||||
int n = size.y;
|
||||
int mini_batch_element_count = n / n_mini_batch;
|
||||
|
||||
Ref<MLPPVector> row_tmp;
|
||||
row_tmp.instance();
|
||||
row_tmp->resize(size.x);
|
||||
|
||||
CreateMiniBatchMVBatch ret;
|
||||
|
||||
for (int i = 0; i < n_mini_batch; i++) {
|
||||
int mini_batch_start_offset = n_mini_batch * i;
|
||||
Ref<MLPPMatrix> current_input_set;
|
||||
current_input_set.instance();
|
||||
current_input_set->resize(Size2i(size.x, mini_batch_element_count));
|
||||
|
||||
Ref<MLPPVector> current_output_set;
|
||||
current_output_set.instance();
|
||||
current_output_set->resize(mini_batch_element_count);
|
||||
|
||||
for (int j = 0; j < mini_batch_element_count; j++) {
|
||||
int main_indx = mini_batch_start_offset + j;
|
||||
|
||||
input_set->get_row_into_mlpp_vector(main_indx, row_tmp);
|
||||
current_input_set->set_row_mlpp_vector(j, row_tmp);
|
||||
|
||||
current_output_set->set_element(j, output_set->get_element(j));
|
||||
}
|
||||
|
||||
ret.input_sets.push_back(current_input_set);
|
||||
ret.output_sets.push_back(current_output_set);
|
||||
}
|
||||
|
||||
/* Don't think this can ever happen, todo double check
|
||||
if (real_t(n) / real_t(n_mini_batch) - int(n / n_mini_batch) != 0) {
|
||||
for (int i = 0; i < n - n / n_mini_batch * n_mini_batch; i++) {
|
||||
inputMiniBatches[n_mini_batch - 1].push_back(inputSet[n / n_mini_batch * n_mini_batch + i]);
|
||||
outputMiniBatches[n_mini_batch - 1].push_back(outputSet[n / n_mini_batch * n_mini_batch + i]);
|
||||
}
|
||||
}
|
||||
*/
|
||||
|
||||
return ret;
|
||||
}
|
||||
MLPPUtilitiesOld::CreateMiniBatchMMBatch MLPPUtilitiesOld::create_mini_batchesmm(const Ref<MLPPMatrix> &input_set, const Ref<MLPPMatrix> &output_set, int n_mini_batch) {
|
||||
Size2i input_set_size = input_set->size();
|
||||
Size2i output_set_size = output_set->size();
|
||||
|
||||
int n = input_set_size.y;
|
||||
int mini_batch_element_count = n / n_mini_batch;
|
||||
|
||||
Ref<MLPPVector> input_row_tmp;
|
||||
input_row_tmp.instance();
|
||||
input_row_tmp->resize(input_set_size.x);
|
||||
|
||||
Ref<MLPPVector> output_row_tmp;
|
||||
output_row_tmp.instance();
|
||||
output_row_tmp->resize(output_set_size.x);
|
||||
|
||||
CreateMiniBatchMMBatch ret;
|
||||
|
||||
for (int i = 0; i < n_mini_batch; i++) {
|
||||
int mini_batch_start_offset = n_mini_batch * i;
|
||||
Ref<MLPPMatrix> current_input_set;
|
||||
current_input_set.instance();
|
||||
current_input_set->resize(Size2i(input_set_size.x, mini_batch_element_count));
|
||||
|
||||
Ref<MLPPMatrix> current_output_set;
|
||||
current_output_set.instance();
|
||||
current_output_set->resize(Size2i(output_set_size.x, mini_batch_element_count));
|
||||
|
||||
for (int j = 0; j < mini_batch_element_count; j++) {
|
||||
int main_indx = mini_batch_start_offset + j;
|
||||
|
||||
input_set->get_row_into_mlpp_vector(main_indx, input_row_tmp);
|
||||
current_input_set->set_row_mlpp_vector(j, input_row_tmp);
|
||||
|
||||
output_set->get_row_into_mlpp_vector(main_indx, output_row_tmp);
|
||||
current_output_set->set_row_mlpp_vector(j, output_row_tmp);
|
||||
}
|
||||
|
||||
ret.input_sets.push_back(current_input_set);
|
||||
ret.output_sets.push_back(current_output_set);
|
||||
}
|
||||
|
||||
/* Don't think this can ever happen, todo double check
|
||||
if (real_t(n) / real_t(n_mini_batch) - int(n / n_mini_batch) != 0) {
|
||||
for (int i = 0; i < n - n / n_mini_batch * n_mini_batch; i++) {
|
||||
inputMiniBatches[n_mini_batch - 1].push_back(inputSet[n / n_mini_batch * n_mini_batch + i]);
|
||||
}
|
||||
}
|
||||
*/
|
||||
|
||||
return ret;
|
||||
}
|
||||
|
||||
Array MLPPUtilitiesOld::create_mini_batchesm_bind(const Ref<MLPPMatrix> &input_set, int n_mini_batch) {
|
||||
Vector<Ref<MLPPMatrix>> batches = create_mini_batchesm(input_set, n_mini_batch);
|
||||
|
||||
Array ret;
|
||||
|
||||
for (int i = 0; i < batches.size(); ++i) {
|
||||
ret.push_back(batches[i].get_ref_ptr());
|
||||
}
|
||||
|
||||
return ret;
|
||||
}
|
||||
Array MLPPUtilitiesOld::create_mini_batchesmv_bind(const Ref<MLPPMatrix> &input_set, const Ref<MLPPVector> &output_set, int n_mini_batch) {
|
||||
CreateMiniBatchMVBatch batches = create_mini_batchesmv(input_set, output_set, n_mini_batch);
|
||||
|
||||
Array inputs;
|
||||
Array outputs;
|
||||
|
||||
for (int i = 0; i < batches.input_sets.size(); ++i) {
|
||||
inputs.push_back(batches.input_sets[i].get_ref_ptr());
|
||||
outputs.push_back(batches.output_sets[i].get_ref_ptr());
|
||||
}
|
||||
|
||||
Array ret;
|
||||
|
||||
ret.push_back(inputs);
|
||||
ret.push_back(outputs);
|
||||
|
||||
return ret;
|
||||
}
|
||||
Array MLPPUtilitiesOld::create_mini_batchesmm_bind(const Ref<MLPPMatrix> &input_set, const Ref<MLPPMatrix> &output_set, int n_mini_batch) {
|
||||
CreateMiniBatchMMBatch batches = create_mini_batchesmm(input_set, output_set, n_mini_batch);
|
||||
|
||||
Array inputs;
|
||||
Array outputs;
|
||||
|
||||
for (int i = 0; i < batches.input_sets.size(); ++i) {
|
||||
inputs.push_back(batches.input_sets[i].get_ref_ptr());
|
||||
outputs.push_back(batches.output_sets[i].get_ref_ptr());
|
||||
}
|
||||
|
||||
Array ret;
|
||||
|
||||
ret.push_back(inputs);
|
||||
ret.push_back(outputs);
|
||||
|
||||
return ret;
|
||||
}
|
||||
|
||||
std::tuple<real_t, real_t, real_t, real_t> MLPPUtilitiesOld::TF_PN(std::vector<real_t> y_hat, std::vector<real_t> y) {
|
||||
real_t TP = 0;
|
||||
real_t FP = 0;
|
||||
|
||||
@ -8,22 +8,13 @@
|
||||
// Created by Marc Melikyan on 1/16/21.
|
||||
//
|
||||
|
||||
#include "core/containers/vector.h"
|
||||
#include "core/math/math_defs.h"
|
||||
#include "core/string/ustring.h"
|
||||
#include "core/variant/variant.h"
|
||||
|
||||
#include "core/object/reference.h"
|
||||
|
||||
#include "../lin_alg/mlpp_matrix.h"
|
||||
#include "../lin_alg/mlpp_vector.h"
|
||||
|
||||
#include <string>
|
||||
#include <tuple>
|
||||
#include <vector>
|
||||
|
||||
class MLPPUtilitiesOld {
|
||||
|
||||
public:
|
||||
// Weight Init
|
||||
static std::vector<real_t> weightInitialization(int n, std::string type = "Default");
|
||||
@ -32,30 +23,10 @@ public:
|
||||
static std::vector<std::vector<real_t>> weightInitialization(int n, int m, std::string type = "Default");
|
||||
static std::vector<real_t> biasInitialization(int n);
|
||||
|
||||
enum WeightDistributionType {
|
||||
WEIGHT_DISTRIBUTION_TYPE_DEFAULT = 0,
|
||||
WEIGHT_DISTRIBUTION_TYPE_XAVIER_NORMAL,
|
||||
WEIGHT_DISTRIBUTION_TYPE_XAVIER_UNIFORM,
|
||||
WEIGHT_DISTRIBUTION_TYPE_HE_NORMAL,
|
||||
WEIGHT_DISTRIBUTION_TYPE_HE_UNIFORM,
|
||||
WEIGHT_DISTRIBUTION_TYPE_LE_CUN_NORMAL,
|
||||
WEIGHT_DISTRIBUTION_TYPE_LE_CUN_UNIFORM,
|
||||
WEIGHT_DISTRIBUTION_TYPE_UNIFORM,
|
||||
};
|
||||
|
||||
void weight_initializationv(Ref<MLPPVector> weights, WeightDistributionType type = WEIGHT_DISTRIBUTION_TYPE_DEFAULT);
|
||||
void weight_initializationm(Ref<MLPPMatrix> weights, WeightDistributionType type = WEIGHT_DISTRIBUTION_TYPE_DEFAULT);
|
||||
real_t bias_initializationr();
|
||||
void bias_initializationv(Ref<MLPPVector> z);
|
||||
|
||||
// Cost/Performance related Functions
|
||||
real_t performance(std::vector<real_t> y_hat, std::vector<real_t> y);
|
||||
real_t performance(std::vector<std::vector<real_t>> y_hat, std::vector<std::vector<real_t>> y);
|
||||
|
||||
real_t performance_vec(const Ref<MLPPVector> &y_hat, const Ref<MLPPVector> &output_set);
|
||||
real_t performance_mat(const Ref<MLPPMatrix> &y_hat, const Ref<MLPPMatrix> &y);
|
||||
real_t performance_pool_int_array_vec(PoolIntArray y_hat, const Ref<MLPPVector> &output_set);
|
||||
|
||||
// Parameter Saving Functions
|
||||
void saveParameters(std::string fileName, std::vector<real_t> weights, real_t bias, bool app = false, int layer = -1);
|
||||
void saveParameters(std::string fileName, std::vector<real_t> weights, std::vector<real_t> initial, real_t bias, bool app = false, int layer = -1);
|
||||
@ -66,42 +37,18 @@ public:
|
||||
static void UI(std::vector<real_t> weights, std::vector<real_t> initial, real_t bias);
|
||||
static void UI(std::vector<std::vector<real_t>> weights, std::vector<real_t> bias);
|
||||
|
||||
static void print_ui_vb(Ref<MLPPVector> weights, real_t bias);
|
||||
static void print_ui_vib(Ref<MLPPVector> weights, Ref<MLPPVector> initial, real_t bias);
|
||||
static void print_ui_mb(Ref<MLPPMatrix> weights, Ref<MLPPVector> bias);
|
||||
|
||||
static void CostInfo(int epoch, real_t cost_prev, real_t Cost);
|
||||
static void cost_info(int epoch, real_t cost_prev, real_t cost);
|
||||
|
||||
static std::vector<std::vector<std::vector<real_t>>> createMiniBatches(std::vector<std::vector<real_t>> inputSet, int n_mini_batch);
|
||||
static std::tuple<std::vector<std::vector<std::vector<real_t>>>, std::vector<std::vector<real_t>>> createMiniBatches(std::vector<std::vector<real_t>> inputSet, std::vector<real_t> outputSet, int n_mini_batch);
|
||||
static std::tuple<std::vector<std::vector<std::vector<real_t>>>, std::vector<std::vector<std::vector<real_t>>>> createMiniBatches(std::vector<std::vector<real_t>> inputSet, std::vector<std::vector<real_t>> outputSet, int n_mini_batch);
|
||||
|
||||
struct CreateMiniBatchMVBatch {
|
||||
Vector<Ref<MLPPMatrix>> input_sets;
|
||||
Vector<Ref<MLPPVector>> output_sets;
|
||||
};
|
||||
|
||||
struct CreateMiniBatchMMBatch {
|
||||
Vector<Ref<MLPPMatrix>> input_sets;
|
||||
Vector<Ref<MLPPMatrix>> output_sets;
|
||||
};
|
||||
|
||||
static Vector<Ref<MLPPMatrix>> create_mini_batchesm(const Ref<MLPPMatrix> &input_set, int n_mini_batch);
|
||||
static CreateMiniBatchMVBatch create_mini_batchesmv(const Ref<MLPPMatrix> &input_set, const Ref<MLPPVector> &output_set, int n_mini_batch);
|
||||
static CreateMiniBatchMMBatch create_mini_batchesmm(const Ref<MLPPMatrix> &input_set, const Ref<MLPPMatrix> &output_set, int n_mini_batch);
|
||||
|
||||
Array create_mini_batchesm_bind(const Ref<MLPPMatrix> &input_set, int n_mini_batch);
|
||||
Array create_mini_batchesmv_bind(const Ref<MLPPMatrix> &input_set, const Ref<MLPPVector> &output_set, int n_mini_batch);
|
||||
Array create_mini_batchesmm_bind(const Ref<MLPPMatrix> &input_set, const Ref<MLPPMatrix> &output_set, int n_mini_batch);
|
||||
|
||||
// F1 score, Precision/Recall, TP, FP, TN, FN, etc.
|
||||
std::tuple<real_t, real_t, real_t, real_t> TF_PN(std::vector<real_t> y_hat, std::vector<real_t> y); //TF_PN = "True", "False", "Positive", "Negative"
|
||||
real_t recall(std::vector<real_t> y_hat, std::vector<real_t> y);
|
||||
real_t precision(std::vector<real_t> y_hat, std::vector<real_t> y);
|
||||
real_t accuracy(std::vector<real_t> y_hat, std::vector<real_t> y);
|
||||
real_t f1_score(std::vector<real_t> y_hat, std::vector<real_t> y);
|
||||
|
||||
};
|
||||
|
||||
#endif /* Utilities_hpp */
|
||||
|
||||
Loading…
Reference in New Issue
Block a user