Causal Effect Estimation and Optimal Dose Suggestions in Mobile Health

Authors: Liangyu Zhu, Wenbin Lu, Rui Song

ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Simulation studies and an application to the Ohio type 1 diabetes dataset show that our method could provide meaningful insights for dose suggestions with mobile health data. 4. Simulation Studies, 5. Type 1 Diabetes Data Analysis
Researcher Affiliation Academia 1Department of Statistics, North Carolina State University, Raleigh, NC, USA. Correspondence to: Liangyu Zhu <lzhu12@ncsu.edu>.
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code Yes The R code for the simulation can be found in https:// github.com/lz2379/Mhealth. The R code for real data application can be found in https: //github.com/lz2379/Mhealth.
Open Datasets Yes We apply our method to the Ohio type 1 diabetes dataset collected by Marling & Bunescu (2018)... Marling, C. and Bunescu, R. C. The ohiot1dm dataset for blood glucose level prediction. In KHD@ IJCAI, pp. 60 63, 2018.
Dataset Splits No The paper only specifies a training and testing split: "We further take the first 44 days as the training data and the last 10 days as the testing data." There is no explicit mention of a validation split.
Hardware Specification No The paper does not provide any specific details about the hardware used for running its experiments.
Software Dependencies No The paper mentions that the code is in R ("The R code for the simulation", "The R code for real data application") but does not specify any libraries or their version numbers.
Experiment Setup Yes We take σ = 0.5, θ1 = 0.8, θ2 = 0, η1 = 0.2, η2 = 0.2, τ1 = 1, τ2 = 0.5, β0 = 0, β1 = 2 and St = Xt. We use the Gaussian kernel KΛ(s) = (2π) q/2|Λ| 1/2 exp( s T Λs/2), where q = 1 is the dimension of St, and f(St) = St. ... Λ is a q q diagonal matrix with Λj,j = λ2 j. We take λj = 0.305 n 1/3sd(St,j), j = 1, . . . , q.