In this episode of Forecasting Impact, hosts Mahdi Abolghasemi and Mariana Menchero speak with Marco Peixeiro, applied data scientist at Nixtla, about the growing importance of explainability in time series forecasting. Marco shares how his work bridges research and practice, from developing deep learning models in NeuralForecast to writing educational resources that make complex forecasting concepts accessible to all.We discuss how explainability builds trust in complex models, the role of SHAP values in understanding forecasts, and what the future holds for interpretability in forecasting. Marco also gives us a preview of his upcoming book, Time Series Forecasting Using Foundation Models, exploring how foundation models are transforming the forecasting landscape.Marco’s books, including the upcoming Time Series Forecasting Using Foundations Models: https://www.manning.com/authors/marco-peixeiro

Ver todos los episodios