Data and code for "Efficient Recursive Estimation for Right-Censored Data with Iterate Averaging"

Description

This repository serves as a companion to the manuscript titled "Efficient Recursive Estimation for Right-Censored Data with Iterate Averaging" (submitted). It provides the complete empirical foundation and technical implementation required to replicate the findings reported in the study. The dataset and accompanying scripts allow for the full reproduction of: Real-data application described in Section 5.7, utilizing the kidney_transplant.csv dataset. Finite-sample gain-sweep analysis presented in Section 5.1, including the generation of "MSE vs c" figures. Repository Structure /data: Contains the benchmark kidney_transplant.csv dataset and a descriptive data_dictionary.csv. /results: Includes all numerical outputs, such as MSE gain-sweep data, convergence trajectories, and summary statistics. /code: Contains the self-contained Python scripts (.py) used to reproduce experimental results and generate figures, supported by a requirements.txt file for environment management. Authors & Contact Authors: N. Nurmukhamedova and N. Boltaeva (National University of Uzbekistan, Faculty of Applied Mathematics and Intelligent Technologies). ORCID: 0000-0002-4672-1722 (Nurmukhamedova); 0009-0005-8194-0364 (Boltaeva). Inquiries: Please direct technical questions or collaboration requests to n.nurmuhamedova@nuu.uz or boltayevan27@gmail.com. Citation To ensure proper attribution, please cite both this repository and the associated manuscript. [Citation placeholder to be updated upon DOI assignment].

Steps to reproduce

Environment Setup: Ensure you have Python 3.12 installed. Reproduce Real-Data Results (Section 5.7): Execute code/sim_real_data.py. This script processes the data/kidney_transplant.csv dataset and generates the results stored in results/realdata_*.csv. Reproduce Gain-Sweep Results (Section 5.1): Execute code/sim_mse_vs_c.py to generate the MSE data, then use code/make_fig_mse_vs_c.py to render the "MSE vs c" figure. Environment Consistency: Use the provided requirements.txt to ensure all necessary libraries are installed in your Python environment. All simulations are controlled by fixed random seeds for full reproducibility.

Institutions

• National University of Uzbekistan

  Tashkent, Tashkent

Categories

Licence