Volcanic eruptions pose a significant threat to human life and economy. Volcanoes of mafic composition are widespread and display different eruptive behaviours, ranging from effusive to highly explosive. Understanding the deep magmatic dynamics that govern eruptive styles is fundamental but challenging due to the complex variation of magma properties at depth. Petrological studies aim to disentangle this complexity, though they encounter challenges as different deep magmatic processes can produce similar petrological signals. This study combines statistical and experimental approaches to investigate the processes behind petrological variations, particularly focusing on how ascent rates impact the chemistry and texture of erupted products. Experimental findings provide fresh insights into the chemical variability observed during the February-April 2021 eruption of Mt. Etna. Additionally, the use of supervised and unsupervised learning techniques enables us to correlate chemical variations at Mt. Etna with changes in other geophysical and volcanological monitoring parameters.