{"id":13510727,"url":"https://github.com/claudiashi57/dragonnet","last_synced_at":"2025-03-30T16:35:03.618Z","repository":{"id":38352545,"uuid":"190123576","full_name":"claudiashi57/dragonnet","owner":"claudiashi57","description":null,"archived":false,"fork":false,"pushed_at":"2022-04-03T07:23:04.000Z","size":5846,"stargazers_count":236,"open_issues_count":2,"forks_count":50,"subscribers_count":8,"default_branch":"master","last_synced_at":"2024-08-02T02:17:12.719Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/claudiashi57.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2019-06-04T03:37:03.000Z","updated_at":"2024-07-16T08:33:25.000Z","dependencies_parsed_at":"2022-08-09T03:01:41.008Z","dependency_job_id":null,"html_url":"https://github.com/claudiashi57/dragonnet","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/claudiashi57%2Fdragonnet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/claudiashi57%2Fdragonnet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/claudiashi57%2Fdragonnet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/claudiashi57%2Fdragonnet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/claudiashi57","download_url":"https://codeload.github.com/claudiashi57/dragonnet/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":222566739,"owners_count":17004237,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-08-01T02:01:52.048Z","updated_at":"2025-03-30T16:35:03.611Z","avatar_url":"https://github.com/claudiashi57.png","language":"Python","funding_links":[],"categories":["Python","其他_生物医药","others"],"sub_categories":["网络服务_其他"],"readme":"# Introduction\n\nThis repository contains software and data for \"[Adapting Neural Networks for the Estimation of Treatment Effects](https://arxiv.org/pdf/1906.02120.pdf)\".\n\nThe paper describes approaches to estimating causal effects from observational data using neural networks. The high-level idea is to modify standard neural net design and training in order to induce a bias towards accurate estimates.\n\n# Requirements and setup\nYou will need to install tensorflow 1.13, sklearn, numpy 1.15, keras 2.2.4 and, pandas 0.24.1\n\n# Data\n\n1. IHDP\nThis dataset is based on a randomized experiment investigating the effect of home visits by specialists on future cognitive scores. \nIt is generated via the npci package [`https://github.com/vdorie/npci`](https://github.com/vdorie/npci) (setting A)\nFor convenience, we have also uploaded a portion of the simulated data in the dat folder. \nThis can be used for testing the code. \n\n\n2. ACIC\nACIC is a collection of semi-synthetic datasets derived from the linked birth and infant death data (LBIDD)\n- Here is the full dataset description [`https://www.researchgate.net/publication/11523952_Infant_Mortality_Statistics_from_the_1999_Period_Linked_BirthInfant_Death_Data_Set`](https://www.researchgate.net/publication/11523952_Infant_Mortality_Statistics_from_the_1999_Period_Linked_BirthInfant_Death_Data_Set)\n- Here is the GitHub repo associated with the competition  [`https://github.com/IBM-HRL-MLHLS/IBM-Causal-Inference-Benchmarking-Framework/blob/master/data/LBIDD/scaling_params.csv`](https://github.com/IBM-HRL-MLHLS/IBM-Causal-Inference-Benchmarking-Framework/blob/master/data/LBIDD/scaling_params.csv)\n- For access to the ACIC 2018 competition data: Please see here [`https://www.synapse.org/#!Synapse:syn11294478/wiki/486304`](https://www.synapse.org/#!Synapse:syn11294478/wiki/486304)\n\n# Reproducing neural net training for IHDP experiments\nThe default setting would let you run Dragonnet, TARNET, and NEDnet under targeted regularization and default mode\n\nYou'll run the from `src` code as \n`./experiment/run_ihdp.sh`\nBefore doing this, you'll need to edit `run_ihdp.sh` and change the following:\n`data_base_dir= where you stored the data`\n`output_base_dir=wherer you want the result to be`\n\nIf you only want to run one of the frameworks, delete the rest of the options in `run_ihdp.sh`\n\n# Reproducing neural net training for the ACIC experiment\nSame as above except you run the from `src` code as `./experiment/run_acic.sh`\n\n# Computing the ATE\nAll of the estimators functions are in `semi_parametric_estimation.ate`\n\nTo reproduce the table in the paper: i) get the neural net predictions; ii) update the output file location in `ihdp_ate.py` iii) run `ihdp_ate.py`. The `make_table` function should generate the mean absolute error for each framework. \n\nNote: the default code use all the data for prediction and estimation. If you want to get the in-sample or out-sample error: i) change the `train_test_split` criteria in `ihdp_main.py`; ii) rerun the neural net training; iii) run `ihdp_ate.py` with apporiate in-sample data and out-sample data. \n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fclaudiashi57%2Fdragonnet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fclaudiashi57%2Fdragonnet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fclaudiashi57%2Fdragonnet/lists"}