Imbalances in covariates between treatment groups are frequent in observational studies and can lead to biased comparisons. Various adjustment methods can be employed to correct these biases in the context of multi-level treatments ( > 2). Analytical challenges, such as positivity violations and incorrect model specification due to unknown functional relationships between covariates and treatment or outcome, may affect their ability to yield unbiased results. Such challenges were expected in a comparison of fire-suppression interventions for preventing fire growth. We identified the overlap weights, augmented overlap weights, bias-corrected matching and targeted maximum likelihood as methods with the best potential to address those challenges. A simple variance estimator for the overlap weight estimators that can naturally be combined with machine learning is proposed. In a simulation study, we investigated the performance of these methods as well as those of simpler alternatives. Adjustment methods that included an outcome modeling component performed better than those that focused on the treatment mechanism in our simulations. Additionally, machine learning implementation was observed to efficiently compensate for the unknown model specification for the former methods, but not the latter. Based on these results, we compared the effectiveness of fire-suppression interventions using the augmented overlap weight estimator.

In this paper we propose and assess the accuracy of new fuel consumption estimation models for vehicle routing. Based on real-world shipping operation data consisting of instantaneous fuel consumption, time-varying speeds observations, and highfrequency traffic, we propose effective methods to estimate fuel consumption. By carrying out nonlinear regression analysis using supervised learning methods, namely Neural Networks, Support Vector Machines, Conditional Inference Trees, and Gradient Boosting Machines, we develop new models that provide better prediction accuracy than classical models. We correctly estimate consumption for time-dependent point-to-point routing under realistic conditions taking into account freight transportation operations during peak hour traffic congestion, stop-and-go driving patterns, idle vehicle states, and the variation of vehicle loads. Our methods provide a more precise alternative to classical regression methods used in the literature, as they are developed for a specific situation (fleet, drivers, geography, etc). Extensive computational experiments under realistic conditions show the effectiveness of the proposed machine learning consumption models, clearly outperforming macroscopic and microscopic consumption models such as the Comprehensive Modal Emissions Model (CMEM) and the Methodology for Estimating air pollutant Emissions from Transport (MEET). Based on sensitivity analyses we show that MEET underestimates real-world consumption by 24.94% and CMEM leads to an overestimation of consumption by 7.57% with optimized parameters. Our best machine learning model (Gradient Boosting Machines) exhibited superior estimation accuracy with a gap of only 1.70%.