Adding YML for NBN, XGB, and UCDavis

model-metadata/CDPH-NBN.yml

@@ -0,0 +1,29 @@
+team_name: "California Department of Public Health"
+team_abbr: "CDPH"
+model_name: "Negative binomial baseline model."
+model_abbr: "NBN"
+model_version: "1.0"
+model_contributors: [
+  {
+    "name": "Hector M. Sanchez C.",
+    "affiliation": "California Department of Public Health",
+    "email": "hector.sanchez-castellanos@cdph.ca.gov"
+  },
+  {
+    "name": "Tomas M. Leon",
+    "affiliation": "California Department of Public Health",
+    "email": "tomas.leon@cdph.ca.gov"
+  }
+]
+website_url: "https://github.com/cdphmodeling/wnvca-2024"
+license: "CC-BY-4.0"
+citation: ""
+team_funding: ""
+designated_model: false
+methods: "Negative binomial model fitted with data from previous year."
+data_inputs: ""
+methods_long: >
+  "A negative binomial distribution was fitted to the number of cases using the number of cases in the same month from the previous year.
+  All counties in the challenge were used in the Bayesian regression, which follows the form: Cases[month, year] ~ Cases[month, year-1]."
+ensemble_of_models: false
+ensemble_of_hub_models: false

model-metadata/CDPH-XGB.yml

@@ -0,0 +1,31 @@
+team_name: "California Department of Public Health"
+team_abbr: "CDPH"
+model_name: "XGBoost cases predictor."
+model_abbr: "XGB"
+model_version: "1.0"
+model_contributors: [
+  {
+    "name": "Hector M. Sanchez C.",
+    "affiliation": "California Department of Public Health",
+    "email": "hector.sanchez-castellanos@cdph.ca.gov"
+  },
+  {
+    "name": "Tomas M. Leon",
+    "affiliation": "California Department of Public Health",
+    "email": "tomas.leon@cdph.ca.gov"
+  }
+]
+website_url: "https://github.com/cdphmodeling/wnvca-2024"
+license: "CC-BY-4.0"
+citation: ""
+team_funding: ""
+designated_model: false
+methods: "XGBoost algorithm used to predict the number of cases in a monthly basis."
+data_inputs: "Historic WNV cases, temperature, precipitation, mosquito counts, dead bird reports."
+methods_long: >
+  "Two predictors were trained upon the data. The first one is a binary classifier to detect if any cases are present, 
+  and returns the probabilities of P(C=0) and P(C>0). The second classifier assesses the probability of cases being in any of the non-zero 
+  classification bins and returns a vector of probabilities P(C=bin). Finally, we evaluate both predictors on each data point and retrieve the full 
+  distribution with: P(C=0), P(C>0)*P(C=bin)."
+ensemble_of_models: false
+ensemble_of_hub_models: false

model-metadata/UCD-GLM.yml

@@ -0,0 +1,27 @@
+team_name: "UC Davis"
+team_abbr: "UCD"
+model_name: ""
+model_abbr: "GLM"
+model_version: "1.0"
+model_contributors: [
+  {
+    "name": "Aynaz Lotfata",
+    "affiliation": "UC Davis",
+    "email": "EMAIL@UCDAVIS.com"
+  },
+  {
+    "name": "Christopher M Barker ",
+    "affiliation": "UC Davis",
+    "email": "EMAIL@UCDAVIS.com"
+  }
+]
+website_url: "https://barkerlab.ucdavis.edu/"
+license: "CC-BY-4.0"
+citation: ""
+team_funding: ""
+designated_model: true
+methods: ""
+data_inputs: ""
+methods_long: "Various datasets containing information on West Nile Virus (WNV) cases, environmental factors, and biological variables are prepared and merged based on county, year, and month. The script then handles missing values and adds lagged climate variables such as temperature, precipitation, and Palmer Drought Severity Index to the dataset. Subsequently, a Negative Binomial Generalized Linear Model (GLM) is fitted to predict WNV cases using the lagged climate variables. Finally, the model is used to make predictions for new data. Additionally, the script mentions a covariate adjustment approach involving the evaluation of RMSE changes when dropping covariates from the model, although the specific implementation of this step is not provided."
+ensemble_of_models: false
+ensemble_of_hub_models: false