{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/estimating-individual-treatment-effect","title":"Estimating individual treatment effect: generalization bounds and algorithms","arxiv_id":"1606.03976","date":"2016-06-13","proceeding":"ICML 2017 8","authors":["Uri Shalit","Fredrik D. Johansson","David Sontag"],"abstract":"There is intense interest in applying machine learning to problems of causal\ninference in fields such as healthcare, economics and education. In particular,\nindividual-level causal inference has important applications such as precision\nmedicine. We give a new theoretical analysis and family of algorithms for\npredicting individual treatment effect (ITE) from observational data, under the\nassumption known as strong ignorability. The algorithms learn a \"balanced\"\nrepresentation such that the induced treated and control distributions look\nsimilar. We give a novel, simple and intuitive generalization-error bound\nshowing that the expected ITE estimation error of a representation is bounded\nby a sum of the standard generalization-error of that representation and the\ndistance between the treated and control distributions induced by the\nrepresentation. We use Integral Probability Metrics to measure distances\nbetween distributions, deriving explicit bounds for the Wasserstein and Maximum\nMean Discrepancy (MMD) distances. Experiments on real and simulated data show\nthe new algorithms match or outperform the state-of-the-art.","url_abs":"http://arxiv.org/abs/1606.03976v5","url_pdf":"http://arxiv.org/pdf/1606.03976v5.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"estimating-individual-treatment-effect","repo_url":"https://github.com/clinicalml/cfrnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"estimating-individual-treatment-effect","repo_url":"https://github.com/oddrose/cfrnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"estimating-individual-treatment-effect","repo_url":"https://github.com/sschrod/bites","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"BSD-2-Clause"}},{"paper_slug":"estimating-individual-treatment-effect","repo_url":"https://github.com/xinshuli2022/cite","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"causal-inference","task_name":"Causal Inference"},{"task_slug":"generalization-bounds","task_name":"Generalization Bounds"},{"task_slug":"heterogeneous-treatment-effect-estimation","task_name":"Heterogeneous Treatment Effect Estimation"}],"methods":[{"method_slug":"causal-inference","method_name":"Causal inference"}],"datasets_introduced":[{"slug":"ihdp","name":"IHDP","full_name":"Infant Health and Development Program"},{"slug":"jobs","name":"Jobs","full_name":""}],"methods_introduced":[],"results":[{"leaderboard":"/sota/causal-inference-on-ihdp","task":"Causal Inference","dataset":"IHDP","model":"Counterfactual Regression + WASS","rank_in_archive_order":5,"of":13,"metrics":{"Average Treatment Effect Error":"0.27"},"uses_additional_data":false},{"leaderboard":"/sota/causal-inference-on-ihdp","task":"Causal Inference","dataset":"IHDP","model":"TARNet","rank_in_archive_order":6,"of":13,"metrics":{"Average Treatment Effect Error":"0.28"},"uses_additional_data":false},{"leaderboard":"/sota/causal-inference-on-ihdp","task":"Causal Inference","dataset":"IHDP","model":"Causal Forest","rank_in_archive_order":8,"of":13,"metrics":{"Average Treatment Effect Error":"0.4"},"uses_additional_data":false},{"leaderboard":"/sota/causal-inference-on-ihdp","task":"Causal Inference","dataset":"IHDP","model":"Balancing Neural Network","rank_in_archive_order":9,"of":13,"metrics":{"Average Treatment Effect Error":"0.42"},"uses_additional_data":false},{"leaderboard":"/sota/causal-inference-on-ihdp","task":"Causal Inference","dataset":"IHDP","model":"k-NN","rank_in_archive_order":11,"of":13,"metrics":{"Average Treatment Effect Error":"0.79"},"uses_additional_data":false},{"leaderboard":"/sota/causal-inference-on-ihdp","task":"Causal Inference","dataset":"IHDP","model":"Balancing Linear Regression","rank_in_archive_order":12,"of":13,"metrics":{"Average Treatment Effect Error":"0.93"},"uses_additional_data":false},{"leaderboard":"/sota/causal-inference-on-ihdp","task":"Causal Inference","dataset":"IHDP","model":"Random Forest","rank_in_archive_order":13,"of":13,"metrics":{"Average Treatment Effect Error":"0.96"},"uses_additional_data":false},{"leaderboard":"/sota/causal-inference-on-jobs","task":"Causal Inference","dataset":"Jobs","model":"CFR MMD","rank_in_archive_order":3,"of":5,"metrics":{"Average Treatment Effect on the Treated Error":"0.08"},"uses_additional_data":false},{"leaderboard":"/sota/causal-inference-on-jobs","task":"Causal Inference","dataset":"Jobs","model":"CFR WASS","rank_in_archive_order":5,"of":5,"metrics":{"Average Treatment Effect on the Treated Error":"0.09"},"uses_additional_data":false},{"leaderboard":"/sota/heterogeneous-treatment-effect-estimation-on","task":"Heterogeneous Treatment Effect Estimation","dataset":"IHDP","model":"TARNet","rank_in_archive_order":2,"of":2,"metrics":{"PEHE":"0.95"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1606.03976","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1606.03976"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/clinicalml/cfrnet","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/sschrod/bites","reach":{"status":"ok","spdx":"BSD-2-Clause"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/oddrose/cfrnet","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/xinshuli2022/cite","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":5},"by_repo_kind":{"listed":{"samples":5,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"8129e8abb094c8f1","entry":"concordance_td","repo":"sschrod/bites","repo_kind":"listed","path":"bites/utils/concordance.py","file_url":"https://github.com/sschrod/bites/blob/HEAD/bites/utils/concordance.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-2-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"8129e8abb094c8f1"}},{"code_sha256_prefix":"82838449bf144f9e","entry":"cox_ph_loss","repo":"sschrod/bites","repo_kind":"listed","path":"bites/utils/loss.py","file_url":"https://github.com/sschrod/bites/blob/HEAD/bites/utils/loss.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-2-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"82838449bf144f9e"}},{"code_sha256_prefix":"8a1945abc2aa367a","entry":"cox_ph_loss_sorted","repo":"sschrod/bites","repo_kind":"listed","path":"bites/utils/loss.py","file_url":"https://github.com/sschrod/bites/blob/HEAD/bites/utils/loss.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-2-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"8a1945abc2aa367a"}},{"code_sha256_prefix":"da646f8263b84ea4","entry":"idx_at_times","repo":"sschrod/bites","repo_kind":"listed","path":"bites/utils/utils.py","file_url":"https://github.com/sschrod/bites/blob/HEAD/bites/utils/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-2-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"da646f8263b84ea4"}},{"code_sha256_prefix":"6e33868d39c90a75","entry":"kaplan_meier","repo":"sschrod/bites","repo_kind":"listed","path":"bites/utils/utils.py","file_url":"https://github.com/sschrod/bites/blob/HEAD/bites/utils/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-2-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"6e33868d39c90a75"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}