Featured Paper: Individual Treatment Effect Prediction for ALS Patients

This paper helps to illustrate the the trend of marrying machine learning with statistics, to combine their strengths: individualization of estimates and proper capture of weak signals on one side, and management & operational guarantees future unknown data on the other side. It focuses on individual effect prediction.

Abstract: A treatment for a complicated disease may be helpful for some but not all patients,
which makes predicting the treatment effect for new patients important yet challenging.
Here we develop a method for predicting the treatment effect based on patient characteristics
and use it for predicting the effect of the only drug (Riluzole) approved for
treating Amyotrophic Lateral Sclerosis (ALS). Our proposed method of model-based random
forests detects similarities in the treatment effect among patients and on this basis
computes personalised models for new patients. The entire procedure focuses on a base
model, which usually contains the treatment indicator as a single covariate and takes the
survival time or a health or treatment success measurement as primary outcome. This
base model is used both to grow the model-based trees within the forest, in which the
patient characteristics that interact with the treatment are split variables, and to compute
the personalised models, in which the similarity measurements enter as weights. We
applied the personalised models using data from several clinical trials for ALS from the
PRO-ACT database. Our results indicate that some ALS patients benefit more from the
drug Riluzole than others. Our method allows shifting from stratified medicine to personalised
medicine and can also be used in assessing the treatment effect for other diseases
studied in a clinical trial.

Full paper link

Featured Paper: Parties, Models, Mobsters: A New Implementation of Model-Based Recursive Partitioning in R

This paper by Zeileis and Hothorn illustrates the ongoing research trend of marrying machine learning with statistics, to combine their strengths: individualization of estimates and proper capture of weak signals on one side, and management & operational guarantees future unknown data on the other side. This particular paper focuses on model-based recursive partitioning.

Abstract: MOB is a generic algorithm for model-based recursive partitioning (Zeileis, Hothorn,
and Hornik 2008). Rather than fitting one global model to a dataset, it estimates local
models on subsets of data that are “learned” by recursively partitioning. It proceeds in the
following way: (1) fit a parametric model to a data set, (2) test for parameter instability
over a set of partitioning variables, (3) if there is some overall parameter instability, split
the model with respect to the variable associated with the highest instability, (4) repeat
the procedure in each of the resulting subsamples. It is discussed how these steps of the
conceptual algorithm are translated into computational tools in an object-oriented manner,
allowing the user to plug in various types of parametric models. For representing the
resulting trees, the R package partykit is employed and extended with generic infrastructure
for recursive partitions where nodes are associated with statistical models. Compared
to the previously available implementation in the party package, the new implementation
supports more inference options, is easier to extend to new models, and provides more
convenience features.

Full paper