|
xgboost
|
基本的模型参数,用于描述助推器。更多...
#include <learner.h>

公共成员函数 | |
| LearnerModelParam ()=default | |
| LearnerModelParam (Context const *ctx, LearnerModelParamLegacy const &user_param, linalg::Vector< float > base_score, ObjInfo t, MultiStrategy multi_strategy) | |
| LearnerModelParam (bst_feature_t n_features, linalg::Vector< float > base_score, std::uint32_t n_groups, bst_target_t n_targets, MultiStrategy multi_strategy) | |
| linalg::VectorView< float const > | BaseScore (Context const *ctx) const |
| linalg::VectorView< float const > | BaseScore (DeviceOrd device) const |
| void | Copy (LearnerModelParam const &that) |
| bool | IsVectorLeaf () const noexcept |
| bst_target_t | OutputLength () const noexcept |
| bst_target_t | LeafLength () const noexcept |
| bool | Initialized () const |
公共属性 | |
| bst_feature_t | num_feature {0} |
| 特征数量。更多... | |
| std::uint32_t | num_output_group {0} |
| 类别或目标数量。更多... | |
| ObjInfo | task {ObjInfo::kRegression} |
| 当前任务,由目标函数决定。更多... | |
| MultiStrategy | multi_strategy {MultiStrategy::kOneOutputPerTree} |
| 构建多目标模型的策略。更多... | |
基本的模型参数,用于描述助推器。
|
默认 |
| xgboost::LearnerModelParam::LearnerModelParam | ( | Context const * | ctx, |
| LearnerModelParamLegacy const & | user_param, | ||
| linalg::Vector< float > | base_score, | ||
| ObjInfo | t, | ||
| MultiStrategy | multi_strategy | ||
| ) |
|
inline |
| linalg::VectorView<float const> xgboost::LearnerModelParam::BaseScore | ( | Context const * | ctx | ) | const |
| linalg::VectorView<float const> xgboost::LearnerModelParam::BaseScore | ( | DeviceOrd | device | ) | const |
| void xgboost::LearnerModelParam::Copy | ( | LearnerModelParam const & | that | ) |
|
inline |
|
inlinenoexcept |
|
inlinenoexcept |
|
inlinenoexcept |
| MultiStrategy xgboost::LearnerModelParam::multi_strategy {MultiStrategy::kOneOutputPerTree} |
构建多目标模型的策略。
| bst_feature_t xgboost::LearnerModelParam::num_feature {0} |
特征数量。
| std::uint32_t xgboost::LearnerModelParam::num_output_group {0} |
类别或目标数量。
| ObjInfo xgboost::LearnerModelParam::task {ObjInfo::kRegression} |
当前任务,由目标函数决定。