{"id":32645704,"url":"https://github.com/ghurault/mbml-eczema","last_synced_at":"2026-06-13T16:08:21.747Z","repository":{"id":196647959,"uuid":"269655328","full_name":"ghurault/mbml-eczema","owner":"ghurault","description":"Mechanistic Bayesian Machine Learning model of eczema dynamic","archived":false,"fork":false,"pushed_at":"2020-08-17T10:42:27.000Z","size":46,"stargazers_count":2,"open_issues_count":0,"forks_count":1,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-10-09T04:18:56.058Z","etag":null,"topics":["bayesian-statistics","eczema","mechanistic-models","probabilistic-models","stan","timeseries-forecasting"],"latest_commit_sha":null,"homepage":null,"language":"R","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"gpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ghurault.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2020-06-05T13:58:37.000Z","updated_at":"2023-07-25T17:09:48.000Z","dependencies_parsed_at":null,"dependency_job_id":"cb300ba7-b206-4e3d-8280-28af13dc1576","html_url":"https://github.com/ghurault/mbml-eczema","commit_stats":null,"previous_names":["ghurault/mbml-eczema"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/ghurault/mbml-eczema","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ghurault%2Fmbml-eczema","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ghurault%2Fmbml-eczema/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ghurault%2Fmbml-eczema/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ghurault%2Fmbml-eczema/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ghurault","download_url":"https://codeload.github.com/ghurault/mbml-eczema/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ghurault%2Fmbml-eczema/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34290503,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-13T02:00:06.617Z","response_time":62,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":["bayesian-statistics","eczema","mechanistic-models","probabilistic-models","stan","timeseries-forecasting"],"created_at":"2025-10-31T04:31:33.258Z","updated_at":"2026-06-13T16:08:21.742Z","avatar_url":"https://github.com/ghurault.png","language":"R","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Personalised prediction of daily eczema severity scores using a mechanistic machine learning model\n\nThis repository contains the code written for the article by [**Hurault et al. (2020), \"Personalised prediction of daily eczema severity scores  using a mechanistic machine learning model\"**](https://doi.org/10.1111/cea.13717), published in Clinical \u0026 Experimental Allergy.\nThe code is written in the R language for statistical computing and the models using the probabilistic programming language [Stan](https://mc-stan.org/).\n\n## File structure\n\nThe `Models` folder contains:\n\n- Our two models\n  - [`BaseModel.stan`](Models/BaseModel.stan): the auto-regressive model based on an Exponentially Modified Gaussian distribution.\n  - [`ExtendedModel.stan`](Models/ExtendedModel.stan): an extension of this model to include measurements present in SWET but not in the Flares dataset.\nThis model notably takes into account the quantity and potency of the treatment use in the determination of the different treatment responsiveness, as well as demographic factors.\n- Two benchmarks (the other benchmarks, the uniform and historical forecasts, are computed in the [`validation.R`](validation.R) script)\n  - [`Autoregression.stan`](Models/Autoregression.stan): an auto-regressive model (our model without Flares triggers).\n  - [`RandomWalk.stan`](Models/RandomWalk.stan): a Gaussian random walk model.\n\nThese models, except the ExtendedModel, can be fitted to the Flares and SWET datasets in [`fitting.R`](fitting.R).\nThe ExtendedModel can be fitted in [`fitting_ext.R`](fitting_ext.R).\n\nForward chaining is implemented in [`validation.R`](validation.R) and these results analysed in [`results_validation.R`](results_validation.R).\n\nUtility functions used within the scripts are available in [`functions.R`](functions.R).\nIn addition, we used functions from Guillem Hurault's personal package, [HuraultMisc](https://github.com/ghurault/HuraultMisc).\n\nThe Flares and SWET datasets are not available according to our data sharing agreement.\nDuring the analysis, these datasets are loaded from a proprietary package `TanakaData` which includes the raw files as well as data processing functions.\nFurther processing is performed by functions written in [`functions_data.R`](functions_data.R).\n\nNonetheless, it is possible to generate fake data from the prior predictive distribution of the different proposed models.\nWe implement this, as well as prior predictive check and fake data check in [`prior_fake_check.R`](prior_fake_check.R).\n\nFinally, [`plots.R`](plots.R) is used to produce several plots present in the paper.\n\n## License\n\nThis open source version of mbml-eczema is licensed under the GPLv3 license, which can be seen in the [LICENSE](LICENSE) file.\n\nA **closed source** version of mbml-eczema is also available without the restrictions of the GPLv3 license with a software usage agreement from Imperial College London.\nFor more information, please contact [Vaibhav Sharma](mailto:v.sharma@imperial.ac.uk).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fghurault%2Fmbml-eczema","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fghurault%2Fmbml-eczema","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fghurault%2Fmbml-eczema/lists"}