Porphyry copper deposits are a rare manifestation of arc magmatism, occurring in restricted spatiotemporal windows within magmatic arcs. Many of the largest deposits are related to the optimal alignment of tectono-magmatic processes, and this magmatic history can be recorded in the geochemistry of related rocks and minerals. The first half of this thesis documents the magmatic evolution of the Quellaveco District, Southern Peru, where 12 Myr of magmatism is recorded by incremental batholith construction and multi-centred porphyry Cu-Mo deposit emplacement. Whole-rock geochemistry and zircon petrochronology record a deepening locus of lower crustal magmatic evolution ca. 2 Myr prior to mineralisation. This delay allowed a sufficiently large volume of hydrous magma to accumulate in the deep crust which was rapidly transferred to the upper crust to form giant porphyry Cu-Mo deposits. Numerical modelling of zircon Eu anomalies reveals that the characteristically suppressed Eu anomaly of zircons from porphyry copper deposits is predominantly caused by inheriting the melt compositional signature of amphibole-dominated and plagioclase-suppressed fractionation. Apatite inclusions in zircon from these rocks indicates that the Quellaveco magmatic system was fluid saturated for the duration of porphyry Cu mineralisation. Numerical modelling suggests that the system was held at high crystallinity over this period which promoted saturation of highly saline and copper-charged instantaneous fluids. The second half of this thesis highlights value of using machine learning algorithms to aid geochemical exploration for porphyry Cu deposits. Predicting magma fertility using geochemistry and supervised learning provides a significant improvement over more traditional bivariate methods such as Sr/Y ratios or zircon Eu anomalies. Deep learning can be used to recognise and classify the strongly oscillatory zoned euhedral zircon crystals associated with porphyry deposits. Overall, these methods are effective in discriminating porphyry Cu fertility but are mostly ineffective at indicating the size or type of porphyry deposit.