The objective of this project is improving substantially our understanding of the acceleration of the Universe by developing novel effective techniques and semi-analytical methods for non-linear cosmological perturbation theory that will be applied to the data of upcoming high precision surveys. The accelerated expansion of the Universe was first detected in 1998 with Type Ia supernova data, and today there is ample evidence for it from various other observations such as the cosmic microwave background radiation (CMB) and the distribution of galaxies at large scales. We call dark energy the mysterious force that drives the acceleration of the Universe and understanding its origin is one of the major problems in cosmology of our time. The high precision of the near-future probes designed for elucidating the nature of dark energy needs to be matched with very accurate theoretical predictions, in particular concerning non-linear effects in cosmological perturbation theory. Developing efficient methods to compute these effects is a crucial requirement for the interpretation of the future dark energy data. I will apply the methods of effective field theories, that have been widely successful in many areas of physics, to develop new tools for describing dark matter and dark energy perturbations in the linear and non-linear regimes. I will combine these tools with state-of-the-art forecasting techniques to determine the prospects for the detection of dark energy parameters from future large scale structure data (specifically from Euclid) in combination with CMB data from Planck.