Accurate forecasting of urban air pollution is essential for protecting public health and guiding mitigation policies. While Deep Learning (DL) and hybrid pipelines dominate recent research, their complexity and limited interpretability hinder operational use. This study investigates whether lightweight additive models -- Facebook Prophet (FBP) and NeuralProphet (NP) -- can deliver competitive forecasts for particulate matter (PM$_{2.5}$, PM$_{10}$) in Beijing, China. Using multi-year pollutant and meteorological data, we applied systematic feature selection (correlation, mutual information, mRMR), leakage-safe scaling, and chronological data splits. Both models were trained with pollutant and precursor regressors, with NP additionally leveraging lagged dependencies. For context, two machine learning baselines (LSTM, LightGBM) and one traditional statistical model (SARIMAX) were also implemented. Performance was evaluated on a 7-day holdout using MAE, RMSE, and $R^2$. Results show that FBP consistently outperformed NP, SARIMAX, and the learning-based baselines, achieving test $R^2$ above 0.94 for both pollutants. These findings demonstrate that interpretable additive models remain competitive with both traditional and complex approaches, offering a practical balance of accuracy, transparency, and ease of deployment.

Groundwater level prediction is an applied time series forecasting task with important social impacts to optimize water management as well as preventing some natural disasters: for instance, floods or severe droughts. Machine learning methods have been reported in the literature to achieve this task, but they are only focused on the forecast of the groundwater level at a single location. A global forecasting method aims at exploiting the groundwater level time series from a wide range of locations to produce predictions at a single place or at several places at a time. Given the recent success of global forecasting methods in prestigious competitions, it is meaningful to assess them on groundwater level prediction and see how they are compared to local methods. In this work, we created a dataset of 1026 groundwater level time series. Each time series is made of daily measurements of groundwater levels and two exogenous variables, rainfall and evapotranspiration. This dataset is made available to the communities for reproducibility and further evaluation. To identify the best configuration to effectively predict groundwater level for the complete set of time series, we compared different predictors including local and global time series forecasting methods. We assessed the impact of exogenous variables. Our result analysis shows that the best predictions are obtained by training a global method on past groundwater levels and rainfall data.

Prophet models are effective, interpretable, and easy to use. But which one is better? In this post we will explore the implementation differences of Prophet and Neural Prophet and run a quick case study. But before we start coding, let's quickly cover some background information, more of which can be found here. Prophet (2017) is the predecessor to NeuralProphet (2020) -- the latter incorporates some autoregressive deep learning.

We introduce NeuralProphet, a successor to Facebook Prophet, which set an industry standard for explainable, scalable, and user-friendly forecasting frameworks. With the proliferation of time series data, explainable forecasting remains a challenging task for business and operational decision making. Hybrid solutions are needed to bridge the gap between interpretable classical methods and scalable deep learning models. We view Prophet as a precursor to such a solution. However, Prophet lacks local context, which is essential for forecasting the near-term future and is challenging to extend due to its Stan backend. NeuralProphet is a hybrid forecasting framework based on PyTorch and trained with standard deep learning methods, making it easy for developers to extend the framework. Local context is introduced with auto-regression and covariate modules, which can be configured as classical linear regression or as Neural Networks. Otherwise, NeuralProphet retains the design philosophy of Prophet and provides the same basic model components. Our results demonstrate that NeuralProphet produces interpretable forecast components of equivalent or superior quality to Prophet on a set of generated time series. NeuralProphet outperforms Prophet on a diverse collection of real-world datasets. For short to medium-term forecasts, NeuralProphet improves forecast accuracy by 55 to 92 percent.