shap
게임 이론적 접근을 통해 머신러닝 모델의 출력을 설명하기
파일 탐색기
최종 버전 다운로드 (.zip)- pr-templates.md
- SKILL.md
- bug-report.yml
- config.yml
- feature-request.yml
- build_gpu.yml
- build_wheels.yml
- codeql-analysis.yml
- gpu_smoke.yml
- run_notebooks.yml
- run_tests.yml
- stale_issues.yml
- test_js.yml
- codecov.yml
- dependabot.yml
- PULL_REQUEST_TEMPLATE.md
- release.yml
- install.ps1
- run_with_env.cmd
- sim_n01549053_4208.jpg
- sim_n01624305_3109.jpg
- sim_n01642257_3703.jpg
- sim_n01687978_7467.jpg
- sim_n01887474_9375.jpg
- sim_n02010272_1613.jpg
- sim_n02011805_6117.jpg
- sim_n02033561_2663.jpg
- sim_n02138441_3552.jpg
- sim_n02190166_6252.jpg
- sim_n02382204_7427.jpg
- sim_n02395694_14311.jpg
- sim_n02968333_22100.jpg
- sim_n03029197_6381.jpg
- sim_n03049924_13.jpg
- sim_n03061345_7223.jpg
- sim_n03118969_4898.jpg
- sim_n03235180_7802.jpg
- sim_n03245889_16148.jpg
- sim_n03268918_3662.jpg
- sim_n03433637_2002.jpg
- sim_n03488887_15199.jpg
- sim_n03518943_2240.jpg
- sim_n03523987_7290.jpg
- sim_n03809603_3531.jpg
- sim_n03822767_7285.jpg
- sim_n04260364_2410.jpg
- sim_n04288533_9519.jpg
- sim_n04404412_26251.jpg
- sim_n04564581_8170.jpg
- sim_n04583212_9954.jpg
- sim_n07560652_7278.jpg
- sim_n07581931_2919.jpg
- sim_n07663899_6767.jpg
- sim_n07697313_11105.jpg
- sim_n07714571_3105.jpg
- sim_n07910656_7115.jpg
- sim_n07930554_10266.jpg
- sim_n09406793_39048.jpg
- sim_n09691729_2568.jpg
- sim_n09706255_4386.jpg
- sim_n09906704_22866.jpg
- sim_n09988703_14533.jpg
- sim_n10374849_811.jpg
- sim_n11508382_9183.jpg
- sim_n11769176_1973.jpg
- sim_n12560282_16028.jpg
- sim_n12709103_15667.jpg
- sim_n12828791_5672.jpg
- sim_n13052670_6077.jpg
- a1a.svmlight
- adult.data
- CommViolPredUnnormalizedData.txt
- imagenet50_224x224.npy
- imagenet50_labels.csv
- imdb_train.txt
- NHANESI_subset_X.csv
- NHANESI_subset_y.csv
- NHANESI_X.csv
- NHANESI_y.csv
- style.css
- layout.html
- california_beeswarm.png
- california_dataset.html
- california_dataset.png
- california_global_bar.png
- california_instance.html
- california_instance.png
- california_scatter.png
- california_waterfall.png
- favicon.ico
- force_plot_matplotlib_rotate.png
- gradient_imagenet_plot.png
- iris_dataset.png
- iris_instance.png
- logo.ai
- logo.pdf
- logo.png
- mnist_image_plot.png
- nhanes1_chol_interaction.png
- nhanes_age_sex_interaction.png
- sentiment_analysis_plot.png
- shap_diagram.png
- shap_header.png
- shap_header.svg
- shap_logo.png
- shap_logo_scratch.ai
- shap_logo_white.png
- simple_iris_dataset.png
- simple_iris_explanation.png
- strawberry_example.png
- Front Page DeepExplainer MNIST Example.html
- Keras LSTM for IMDB Sentiment Classification.html
- PyTorch Deep Explainer MNIST example.html
- Explain an Intermediate Layer of VGG16 on ImageNet (PyTorch).html
- Explain an Intermediate Layer of VGG16 on ImageNet.html
- Sentiment Analysis with Logistic Regression.html
- crime.pickle
- crime.py
- heart.pickle
- decision_plot.html
- dependence_plot.html
- Census income classification with LightGBM.html
- Census income classification with Keras.html
- Census income classification with LightGBM.html
- Census income classification with scikit-learn.html
- Census income classification with XGBoost.html
- ImageNet VGG16 Model with Keras.html
- Iris classification with scikit-learn.html
- League of Legends Win Prediction with XGBoost.html
- NHANES I Survival Model.html
- SHAP_NIPS_2017_Errata.pdf
- February 2018 Talk.pptx
- Federal Reserve 2021 talk.pptx
- INFORMS 2019.pptx
- NIPS 2017 Talk.pptx
- ODSC West talk.pptx
- TMLS 2020 talk.pptx
- nature_bme.bib
- shap_nips.bib
- tree_explainer.bib
- treeshap_arxiv.bib
- bootstrap.min.css
- style.css
- glyphicons-halflings-regular.eot
- glyphicons-halflings-regular.svg
- glyphicons-halflings-regular.ttf
- glyphicons-halflings-regular.woff
- Example1.png
- Example1_nobar.png
- Example2.png
- Example2_nobar.png
- university.png
- bundle.js
- task.js
- utils.js
- backbone-min.js
- bootstrap.min.js
- d3.v3.min.js
- jquery-min.js
- underscore-min.js
- favicon.ico
- ad.html
- closepopup.html
- complete.html
- consent.html
- default.html
- error.html
- exp.html
- thanks-mturksubmit.html
- thanks.html
- .psiturk_history
- Analyse Results.ipynb
- Approve Hits.ipynb
- config.txt
- Debrief.jsx
- eventdata.csv
- human_and_survey.pdf
- human_or_survey.pdf
- human_sum_survey.pdf
- human_xor_survey.pdf
- index.jsx
- Instructions.jsx
- Main.jsx
- package.json
- participants.db
- questiondata.csv
- Questions.jsx
- README.md
- server.log
- trial_data2_3_3_2019.csv
- trial_data_3_3_2019.csv
- trialdata.csv
- webpack.config.js
- sickness_scores.zip
- api.rst
- api_examples.rst
- benchmarks.rst
- conf.py
- contributing.rst
- genomic_examples.rst
- image_examples.rst
- index.rst
- make.bat
- Makefile
- overviews.rst
- release_notes.rst
- requirements-docs.txt
- tabular_examples.rst
- text_examples.rst
- render.spec.jsx.snap
- render.spec.jsx
- AdditiveForceArrayVisualizer.jsx
- AdditiveForceVisualizer.jsx
- color-set.js
- index.jsx
- SimpleListVisualizer.jsx
- .npmignore
- babel.config.js
- developer-docs.md
- index.jsx
- jest.config.js
- Makefile
- package-lock.json
- package.json
- random-explanation.js
- README.md
- test_bundle.js
- webpack.config.js
- __init__.py
- exact_explainer.py
- partition_explainer.py
- permutation_explainer.py
- tabular_masker.py
- Exact.ipynb
- GPUTree.ipynb
- Permutation.ipynb
- custom.ipynb
- bar.ipynb
- beeswarm.ipynb
- decision_plot.ipynb
- heatmap.ipynb
- image.ipynb
- scatter.ipynb
- text.ipynb
- violin.ipynb
- waterfall.ipynb
- migrating-to-new-api.ipynb
- Image Multiclass Classification Benchmark Demo.ipynb
- Benchmark Debug Mode.ipynb
- Benchmark XGBoost explanations.ipynb
- Tabular Prediction Benchmark Demo.ipynb
- Abstractive Summarization Benchmark Demo.ipynb
- Machine Translation Benchmark Demo.ipynb
- Text Emotion Multiclass Classification Benchmark Demo.ipynb
- DeepExplainer Genomics Example.ipynb
- Image Captioning using Azure Cognitive Services.ipynb
- Image Captioning using Open Source.ipynb
- Explain an Intermediate Layer of VGG16 on ImageNet (PyTorch).ipynb
- Explain an Intermediate Layer of VGG16 on ImageNet.ipynb
- Explain MobilenetV2 using the Partition explainer (PyTorch).ipynb
- Explain ResNet50 using the Partition explainer.ipynb
- Front Page DeepExplainer MNIST Example.ipynb
- Image Multi Class.ipynb
- Multi-class ResNet50 on ImageNet (TensorFlow)-checkpoint.ipynb
- Multi-class ResNet50 on ImageNet (TensorFlow).ipynb
- Multi-input Gradient Explainer MNIST Example.ipynb
- PyTorch Deep Explainer MNIST example.ipynb
- An introduction to explainable AI with Shapley values.ipynb
- Be careful when interpreting predictive models in search of causal insights.ipynb
- Explaining quantitative measures of fairness.ipynb
- Explaining a model that uses standardized features.ipynb
- Math behind LinearExplainer with correlation feature perturbation.ipynb
- Sentiment Analysis with Logistic Regression.ipynb
- Census income classification with scikit-learn.ipynb
- Diabetes regression.ipynb
- Iris classification with scikit-learn.ipynb
- Multioutput Regression SHAP.ipynb
- Simple California Demo.ipynb
- Simple Kernel SHAP.ipynb
- Squashing Effect.ipynb
- Census income classification with Keras.ipynb
- cognitive_score.txt
- module_expression.txt
- neuropath.txt
- tweets.csv
- Figure 1 - Simple Inconsistency Example.ipynb
- Figure 3 - User Study.ipynb
- Figure 4 - Supervised Clustering Adult Census Data.ipynb
- Figure 5 - Runtime.ipynb
- Figure 6 - Supervised Clustering R-squared.ipynb
- Figure 7 - Airline Tweet Sentiment Analysis.ipynb
- Figures 8-11 NHANES I Survival Model-Copy1.ipynb
- Figures 8-11 NHANES I Survival Model.ipynb
- perf.pdf
- Performance comparison copy.ipynb
- Performance comparison.ipynb
- Tree SHAP in Python.ipynb
- Basic SHAP Interaction Value Example in XGBoost.ipynb
- Catboost tutorial.ipynb
- Census income classification with LightGBM.ipynb
- Census income classification with XGBoost.ipynb
- Example of loading a custom tree model into SHAP.ipynb
- Explaining a simple OR function.ipynb
- Explaining the Loss of a Model.ipynb
- Fitting a Linear Simulation with XGBoost.ipynb
- Force Plot Colors.ipynb
- Front page example (XGBoost).ipynb
- global_summary.jpg
- League of Legends Win Prediction with XGBoost.ipynb
- NHANES I Survival Model.ipynb
- Perfomance Comparison.ipynb
- Python Version of Tree SHAP.ipynb
- Scatter Density vs. Violin Plot Comparison.ipynb
- Understanding Tree SHAP for Simple Models.ipynb
- Language Modeling Explanation Demo.ipynb
- Explaining a Question Answering Transformers Model.ipynb
- Emotion classification multiclass example.ipynb
- Keras LSTM for IMDB Sentiment Classification.ipynb
- Positive vs. Negative Sentiment Classification.ipynb
- Using custom functions and tokenizers.ipynb
- Abstractive Summarization Explanation Demo.ipynb
- Textual Entailment Explanation Demo.ipynb
- Open Ended GPT2 Text Generation Explanations.ipynb
- Machine Translation Explanations.ipynb
- run_notebooks_timeouts.py
- __init__.py
- _action.py
- _optimizer.py
- __init__.py
- _compute.py
- _explanation_error.py
- _result.py
- _sequential.py
- experiments.py
- framework.py
- measures.py
- methods.py
- metrics.py
- models.py
- plots.py
- _cext.cc
- _cext_gpu.cc
- _cext_gpu.cu
- gpu_treeshap.h
- tree_shap.h
- __init__.py
- clustering_utils.h
- cutils.cpp
- grey_code_utils.h
- kernel_explainer_utils.h
- tabular_utils.h
- __init__.py
- deep_pytorch.py
- deep_tf.py
- deep_utils.py
- __init__.py
- _coefficient.py
- _lime.py
- _maple.py
- _random.py
- _treegain.py
- _ubjson.py
- __init__.py
- _additive.py
- _coalition.py
- _exact.py
- _explainer.py
- _gpu_tree.py
- _gradient.py
- _kernel.py
- _linear.py
- _partition.py
- _permutation.py
- _sampling.py
- _tree.py
- pytree.py
- tf_utils.py
- __init__.py
- _composite.py
- _fixed.py
- _fixed_composite.py
- _image.py
- _masker.py
- _output_composite.py
- _tabular.py
- _text.py
- __init__.py
- _model.py
- _teacher_forcing.py
- _text_generation.py
- _topk_lm.py
- _transformers_pipeline.py
- __init__.py
- _colorconv.py
- _colors.py
- bundle.js
- logoSmallGray.png
- __init__.py
- _bar.py
- _beeswarm.py
- _benchmark.py
- _decision.py
- _embedding.py
- _force.py
- _force_matplotlib.py
- _group_difference.py
- _heatmap.py
- _image.py
- _labels.py
- _monitoring.py
- _partial_dependence.py
- _scatter.py
- _style.py
- _text.py
- _utils.py
- _violin.py
- _waterfall.py
- __init__.py
- _clustering.py
- _exceptions.py
- _general.py
- _legacy.py
- _masked_model.py
- _show_progress.py
- _types.py
- _warnings.py
- image.py
- transformers.py
- __init__.py
- _explanation.py
- _serializable.py
- datasets.py
- links.py
- test_action.py
- test_optimizer.py
- framework.py
- perturbation.py
- test_imports.py
- test_maple.py
- test_treegain.py
- test_ubjson.py
- __init__.py
- common.py
- conftest.py
- test_coalition.py
- test_deep.py
- test_exact.py
- test_explainer.py
- test_gpu_tree.py
- test_gradient.py
- test_kernel.py
- test_linear.py
- test_partition.py
- test_permutation.py
- test_sampling.py
- test_tree.py
- __init__.py
- test_composite.py
- test_custom.py
- test_fixed.py
- test_fixed_composite.py
- test_image.py
- test_masker.py
- test_masker_with_explainers.py
- test_output_composite.py
- test_tabular.py
- test_text.py
- test_teacher_forcing_logits.py
- test_text_generation.py
- test_bar.png
- test_bar_local_feature_importance.png
- test_bar_with_clustering.png
- test_bar_with_cohorts_dict.png
- test_beeswarm.png
- test_beeswarm_no_group_remaining.png
- test_colormaps_red_blue.png
- test_colormaps_red_blue_circle.png
- test_colormaps_red_blue_no_bounds.png
- test_colormaps_red_blue_transparent.png
- test_colormaps_red_transparent_blue.png
- test_colormaps_red_white_blue.png
- test_colormaps_transparent_blue.png
- test_colormaps_transparent_red.png
- test_decision_multioutput.png
- test_decision_plot.png
- test_decision_plot_interactions.png
- test_decision_plot_single_instance.png
- test_force_array_js.png
- test_force_js.png
- test_force_plot_negative_sign.png
- test_force_plot_positive_sign.png
- test_group_difference.png
- test_heatmap.png
- test_heatmap_feature_order.png
- test_image_multi.png
- test_image_multi_labels_per_row_list.png
- test_image_multi_labels_per_row_ndarray.png
- test_image_multi_no_labels.png
- test_image_single.png
- test_scatter_categorical.png
- test_scatter_custom.png
- test_scatter_dotchain.png
- test_scatter_interaction.png
- test_scatter_multiple_cols_overlay.png
- test_scatter_plot_value_input_input0.png
- test_scatter_plot_value_input_input1.png
- test_scatter_plot_value_input_input2.png
- test_scatter_single.png
- test_summary.png
- test_summary_bar_multiclass.png
- test_summary_bar_with_data.png
- test_summary_compact_dot_with_data.png
- test_summary_dot_with_data.png
- test_summary_layered_violin_with_data.png
- test_summary_multi_class.png
- test_summary_multi_class_legend.png
- test_summary_multi_class_legend_decimals.png
- test_summary_multiclass_explanation.png
- test_summary_plot.png
- test_summary_plot_interaction.png
- test_summary_plot_twice.png
- test_summary_violin_regression.png
- test_summary_violin_with_data.png
- test_summary_with_data.png
- test_summary_with_log_scale.png
- test_violin.png
- test_waterfall.png
- test_waterfall_bounds.png
- test_waterfall_custom_style.png
- test_waterfall_legacy.png
- __init__.py
- conftest.py
- test_bar.py
- test_beeswarm.py
- test_benchmark.py
- test_colors.py
- test_decision.py
- test_dependence.py
- test_dependence_string_features.py
- test_force.py
- test_force_js.py
- test_group_difference.py
- test_heatmap.py
- test_image.py
- test_scatter.py
- test_style.py
- test_summary.py
- test_text.py
- test_utils.py
- test_violin.py
- test_waterfall.py
- test_coalition_test_tabular_coalition_multiple_output_baseline.npz
- test_coalition_test_tabular_coalition_single_output_baseline.npz
- test_exact_test_interactions_baseline.npz
- test_exact_test_serialization_baseline.npz
- test_exact_test_serialization_custom_model_save_baseline.npz
- test_exact_test_serialization_no_model_or_masker_baseline.npz
- test_exact_test_serialization_no_model_or_masker_reduced_baseline.npz
- test_exact_test_tabular_multi_output_auto_masker_baseline.npz
- test_exact_test_tabular_multi_output_independent_masker_baseline.npz
- test_exact_test_tabular_multi_output_partition_masker_baseline.npz
- test_exact_test_tabular_single_output_auto_masker_baseline.npz
- test_exact_test_tabular_single_output_auto_masker_minimal_baseline.npz
- test_exact_test_tabular_single_output_auto_masker_single_value_baseline.npz
- test_exact_test_tabular_single_output_independent_masker_baseline.npz
- test_exact_test_tabular_single_output_partition_masker_baseline.npz
- test_kernel_test_serialization_baseline.npz
- test_partition_test_serialization_baseline.npz
- test_partition_test_serialization_custom_model_save_baseline.npz
- test_partition_test_serialization_no_model_or_masker_baseline.npz
- test_partition_test_tabular_multi_output_baseline.npz
- test_partition_test_tabular_single_output_baseline.npz
- test_partition_test_translation_algorithm_arg_baseline.npz
- test_partition_test_translation_auto_baseline.npz
- test_partition_test_translation_baseline.npz
- test_permutation_test_tabular_single_output_auto_masker_baseline.npz
- test_clustering.py
- test_general.py
- test_masked_model.py
- conftest.py
- datasets_to_cache.py
- gpu_tree_tests.ipynb
- README.md
- test_datasets.py
- test_explanation.py
- test_install.py
- .git-blame-ignore-revs
- .gitattributes
- .gitignore
- .pre-commit-config.yaml
- .readthedocs.yml
- asv.conf.json
- CMakeLists.txt
- CONTRIBUTING.md
- LICENSE
- pyproject.toml
- README.md
# 설치 가이드
1. 코드 내려받기
git clone https://github.com/shap/shap
깃허브에서 프로젝트 코드 전체를 내 컴퓨터로 내려받습니다.
cd shap
방금 내려받은 프로젝트 폴더 안으로 이동합니다.
2. CMake
보통 추천사전 준비물
mkdir build && cd build
빌드 결과물을 담을 폴더를 만들고 그 안으로 이동합니다.
cmake ..
소스코드를 분석해 빌드 설정 파일을 생성합니다 (build 폴더 안에서 실행해야 함).
make
생성된 빌드 설정을 바탕으로 실제 컴파일을 진행해 실행 파일을 만듭니다.
build 폴더 안에 실행 파일이 생성됐는지 확인하고, 직접 실행해보세요 (예: ./build/앱이름).
3. Node.js
쉬움사전 준비물
cd javascript
이 프로젝트의 관련 파일이 하위 폴더 안에 있어서, 먼저 그 폴더로 이동합니다.
npm install
package.json에 명시된 라이브러리들을 내려받아 설치합니다.
npm start
개발/실행 서버를 켭니다.
명령어 실행 후 터미널에 나타나는 주소(보통 http://localhost:3000 형태)를 브라우저에서 열어보세요.
4. Python
쉬움사전 준비물
pip install .
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
jupyter notebook
브라우저에서 노트북(.ipynb) 파일들을 열람하고 실행할 수 있는 Jupyter 화면을 켭니다.
에러 메시지 없이 실행되고 터미널에 안내 문구가 출력되면 정상입니다.
5. Make
보통사전 준비물
- Git GitHub에서 프로젝트 코드를 내려받으려면 필요합니다.
- Make Linux/macOS는 보통 기본 설치되어 있습니다. Windows는 별도 설치(예: MSYS2, WSL)가 필요합니다.
cd javascript
이 프로젝트의 관련 파일이 하위 폴더 안에 있어서, 먼저 그 폴더로 이동합니다.
make
생성된 빌드 설정을 바탕으로 실제 컴파일을 진행해 실행 파일을 만듭니다.
에러 없이 끝나면 성공입니다. 생성된 실행 파일을 직접 실행해보세요.
// repository documentation
Was this content helpful?
(0 ratings)