A Ride in Targeted Learning Territory
2022-10-20
Welcome
This is either the website or the text called “A Ride in Targeted Learning Territory”. In the former case, the text can be dowloaded by clicking on the dedicated button in the top part of the webpage. In the latter case, the website can be browsed at https://achambaz.github.io/tlride/.
Long Mendocino Drive (detail, Liana Steinmetz)
Organization
The text takes the form of a series of brief sections. The main sections
combine theoretical and computational developments. The code is written in the
programming language R
(R Core Team 2019). R
is widely
used among statisticians and data scientists to develop statistical software
and data analysis.
Regularly, a section is inserted that proposes exercises. Each such section is indicated by the ⚙ symbol. The symbol ☡ also indicates those sections of which the content is more involved.
Overview
After a short introduction, we present the reproducible experiment that will play a central role throughout the text. Then, we introduce the main parameter of interest. We comment upon some of its properties that are useful from a statistical perspective. This paves the way to the presentation of several estimators that are increasingly more powerful statistically. The discussion takes us into targeted learning territory.
Audience
The text might be of interest to students in statistics and machine learning. It might also serve as a gentle introduction to targeted learning in both its theoretical and computational aspects before delving deeper into the literature (Laan and Rose 2011), (Laan and Rose 2018).
The text was presented at the Journées d’Étude en Statistique 2018 held in Fréjus (France) in October 2018 and at the First Summer School in Statistics and Data Science for Young Researchers from French-speaking Africa held at AIMS Senegal (MBour, Senegal) in July 2019. The text and a shorter version (in French) will be published soon (Benkeser and Chambaz 2020a), (Benkeser and Chambaz 2020b).
Technical details
The text is written in RMarkdown
with bookdown
. Assuming that
the knitr
package is installed, you can
retrieve all the R
code by cloning this github
repository then running
::purl("abcd.Rmd") knitr