We investigate the descriptions for the observables of inflationary models, in particular, the spectral index of curvature perturbations, the tensor-to-scalar ratio, and the running of the spectral index, in the framework of perfect fluid models and $F(R)$ gravity theories through the reconstruction methods. Furthermore, the perfect fluid and $F(R)$ gravity descriptions of inflation are compared with the recent cosmological observations such as the Planck satellite and BICEP2 experiment. It is demonstrated with explicit examples that perfect fluid may lead to the inflationary universe consistent with the Planck data. It is also shown that several $F(R)$ gravity models, especially, a power-law model gives the best fit values compatible with the spectral index and tensor-to-scalar ratio within the allowed ranges suggested by the Planck and BICEP2 results.

The reconstruction of a (non)canonical scalar field Lagrangian from the dark energy Equation of State (EoS) parameter is studied, where it is shown that any EoS parametrization can be well reconstructed in terms of scalar fields. Several examples of EoS parameters are studied and the particular scalar field Lagrangian is reconstructed. Then, we propose some new parametrizations that may present a (fast) transition to a phantom dark energy EoS (where $w_{DE}<-1$) and the scalar field Lagrangian is also reconstructed numerically. Furthermore, the properties of these parametrizations of the dark energy EoS are studied by using supernovae Ia data (HST Cluster Supernova Survey) combined with Standard Ruler datasets [Cosmic Microwave Background (CMB) and Baryon Acoustic Oscillations (BAO)] and its comparison with the $\Lambda$CDM model is analyzed. Then, the best fit of the models is obtained, which provides some information about whether a phantom transition may be supported by the observations. In this regard, the crossing of the phantom barrier is allowed statistically but the occurrence of a future singularity seems unlikely.

We construct an $F(R)$ gravity theory corresponding to the Weyl invariant two scalar field theory. We investigate whether such $F(R)$ gravity can have the antigravity regions where the Weyl curvature invariant does not diverge at the Big Bang and Big Crunch singularities. It is revealed that the divergence cannot be evaded completely but can be much milder than that in the original Weyl invariant two scalar field theory.

We discuss conformal issues of pure and extended teleparallel gravity. In particular, we present formulations of conformal transformation in teleparallel gravity. Furthermore, we propose conformal scalar and gauge field theories in teleparallel gravity and study conformal torsion gravity. We explicitly demonstrate that a power-law acceleration (including the $\Lambda$CDM universe) as well as the de Sitter expansion of the universe can be realized in extended teleparallel gravity with a conformal scalar field, and that pure teleparallel gravity with a conformal scalar field can lead to the $\Lambda$CDM model. It is also shown that there can exist the de Sitter solution in conformal torsion gravity.

We explore the possibility of further gravitational wave modes in $F(T)$ gravity, where $T$ is the torsion scalar in teleparallelism. It is explicitly demonstrated that gravitational wave modes in $F(T)$ gravity are equivalent to those in General Relativity. This result is achieved by calculating the Minkowskian limit for a class of analytic function of $F(T)$. This consequence is also confirmed by the preservative analysis around the flat background in the weak field limit with the scalar-tensor representation of $F(T)$ gravity.

We propose general f(R,T,R_{\mu\nu}T^{\mu\nu}) theory as generalization of covariant Horava-like gravity with dynamical Lorentz symmetry breaking. FRLW cosmological dynamics for several versions of such theory is considered. The reconstruction of the above action is explicitly done, including the numerical reconstruction for the occurrence of \Lambda CDM universe. De Sitter universe solutions in the presence of non-constant fluid are also presented. The problem of matter instability in f(R,T,R_{\mu\nu}T^{\mu\nu}) gravity is discussed.

In the context of f(R,T) theories of gravity, we study the evolution of scalar cosmological perturbations in the metric formalism. According to restrictions on the background evolution, a specific model within these theories is assumed in order to guarantee the standard continuity equation. Using a completely general procedure, we find the complete set of differential equations for the matter density perturbations. In the case of sub-Hubble modes, the density contrast evolution reduces to a second-order equation. We show that for well-motivated f(R,T) Lagrangians the quasistatic approximation yields to very different results from the ones derived in the frame of the concordance {\Lambda}CDM model constraining severely the viability of such theories.

Modified f(R) gravity is one of the most promising candidates for dark energy, and even for the unification of the whole cosmological evolution, including the inflationary phase. Within this class of theories, the so-called viable modified gravities represent realistic theories that are capable of reproducing late-time acceleration, and satisfy strong constraints at local scales, where General Relativity is recovered. The present manuscript deals with the analysis of the cosmological evolution for some of these models, which indicates that the evolution may enter into a phantom phase, but the behavior may be asymptotically stable. Furthermore, the scalar-tensor equivalence of f(R) gravity is considered, which provides useful information about the possibility of the occurrence of a future singularity. The so-called Little Rip and Pseudo-Rip are also studied in the framework of this class of modified gravities.

One of the so-called viable modified gravities is analyzed. This kind of gravity theories are characterized by a well behavior at local scales, where General Relativity is recovered, while the modified terms become important at the cosmological level, where the late-time accelerating era is reproduced, and even the inflationary phase. In the present work, the future cosmological evolution for one of these models is studied. A transition to the phantom phase is observed. Furthermore, the scalar-tensor equivalence of f(R) gravity is also considered, which provides important information concerning this kind of models.

Modified gravity is one of the most promising candidates for explaining the current accelerating expansion of the Universe, and even its unification with the inflationary epoch. Nevertheless, the wide range of models capable to explain the phenomena of dark energy, imposes that current research focuses on a more precise study of the possible effects of modified gravity may have on both cosmological and local levels. In this paper, we focus on the analysis of a type of modified gravity, the so-called f(R,G) gravity and we perform a deep analysis on the stability of important cosmological solutions. This not only can help to constrain the form of the gravitational action, but also facilitate a better understanding of the behavior of the perturbations in this class of higher order theories of gravity, which will lead to a more precise analysis of the full spectrum of cosmological perturbations in future.

The Ekpyrotic scenario is studied in the context of some extensions of Ho\v{r}ava-Lifshitz gravity. Some particular solutions that lead to cyclic Hubble parameters are analyzed, where the corresponding gravitational actions are reconstructed by using several techniques and auxiliary fields. Its comparison with standard $F(R)$ gravity is performed. In addition, the so-called Little Rip, a stage of the universe evolution when some bounded systems may be dissolute, is also studied in this frame of theories.

The use of Lagrange multipliers in the context of quintessence/phantom scalar fields allows to constrain the behavior of the scalar field, which provides a powerful tool, not only for the reconstruction of cosmological solutions but also for the study of some problems in cosmology and gravitational physics. In the present paper, we focus on the reconstruction of cosmological solutions capable of controlling the cosmological constant value by imposing a constraint on the scalar field, providing a relaxation mechanism of the value of the cosmological constant. The formalism is also extended to the study of phantom scalar fields with a future singularity and their conformal transformation to the Jordan frame, where a type of modified gravity, constrained by the Lagrange multiplier, is obtained.

The behavior of cosmological evolution is studied in the case when a phantom fluid, that contributes to the universe accelerated expansion, is introduced in an F (R) model. At the early stages of the universe history, the dark fluid is seen to give rise to a deceleration of its expansion. For t close to present time it works as an additional contribution to the effective cosmological constant and, later, it produces the transition to a phantom era, which could actually be taking place right now in some regions of the cosmos, and might have observable consequences. For t close to the Rip time, the universe becomes completely dominated by the dark fluid, whose equation of state is phantom-like at that time. Our model, which is able to reproduce the dark energy period quite precisely, may still be modified in such a way that the epoch dominated by an effective cosmological constant, produced by the F (R) term and by the dark fluid contribution, becomes significantly shorter, what happens when a matter term is included. The dark fluid with phantom behavior gives rise to a super-accelerated phase, as compared with the case where just the viable F (R) term contributes. It is also seen explicitly that an F (R) theory can be constructed from a phantom model in a scalar-tensor theory, in which the scalar field does not behave as phantom (in the latter case the action for F (R) would be complex). Promising F (R) models which can cross the phantom divide in a convenient way are constructed explicitly.

At the present paper, it is studied cosmological solutions and its stability in the frame of F(R) Horava-Lifshitz gravity. The perturbations around general spatially flat FRW solutions are analyzed and it is showed that the stability of those solutions will depend on the kind of theory, i.e. on the form of the action F(R), as well as on the parameters contained in any Horava-Lifshitz theory due to the breaking of Lorentz invariance. The (in)stability of a given cosmic solution can restrict the models and gives new observational predictions, and can give a natural explanation on the end of inflation and radiation/matter phases. An explicit example of F(R) is studied, and it is showed that the instability can produce the transition between the different epochs of the Universe history.

The present thesis is focused on the study of FLRW cosmology in modified gravities and with scalar fields. The mystery of dark energy has made that the last decade, many efforts in theoretical physics have been focused on the explanation of the current acceleration of the Universe expansion. On the other hand, an early accelerated epoch, known as inflation, is also required in the standard cosmological model in order to fit the observational data. Here, we study the possibility to explain both effects under the same mechanism, by means of scalar fields or modified gravity. In this sense, F(R) and Gauss-Bonnet gravities are studied. This analysis is also extended to Ho\v{r}ava-Lifshitz gravity, a theory that seems to be power counting renormalizable, in spite of it contains other serious problems that are also discussed in the thesis.

In the context of the so-called Gauss-Bonnet gravity, where the gravitational action includes function of the Gauss-Bonnet invariant, we study cosmological solutions, especially the well-known $\Lambda$CDM model. It is shown that the dark energy contribution and even the inflationary epoch can be explained in the frame of this kind of theories with no need of any other kind of component. Other cosmological solutions are constructed and the rich properties that this kind of theories provide are explored.

At the present work, it is studied the extension of F (R) gravities to the new recently proposed theory of gravity, the so-called Horava-Lifshitz gravity, which provides a way to make the theory power counting renormalizable by breaking Lorentz invariance. It is showed that dark energy can be well explained in the frame of this extension, just in terms of gravity. It is also explored the possibility to unify inflation and late-time acceleration under the same mechanism, providing a natural explanation the accelerated expansion.

We derive a Cardy-Verlinde-like formula which relates the entropy of the closed FRW universe to its energy, and Casimir energy, for a multicomponent coupled fluid. The generalized fluid obeys an inhomogeneous equation of state. A viscous dark fluid is included, and we include also modified gravity using its fluid representation. It is demonstrated how such an expression reduces to the standard Cardy-Verlinde formula corresponding to the 2d CFT entropy only in some special cases. The dynamical entropy bound for a closed FRW universe with dark components is obtained. The universality of the dynamical entropy bound near a future singularity (of all four types), as well as near the Big Bang singularity, is investigated. It is demonstrated that except from some special cases of Type II and Type IV singularities the dynamical entropy bound is violated near the singularity even if quantum effects are taken into account. The dynamical entropy bound seems to be universal for the case of a regular universe, including the asymptotic de Sitter one.

The role of dynamical cosmological Casimir effect to phantom (constant $w$) and oscillating Universe is discussed. It is shown explicitly that its role is not essential near to Big Rip singularity. However, the account of Casimir fluid makes the scale factor approach to Rip time to be faster. Rip time itself maybe changed too.

An occurrence of an oscillating Universe is showed using an inhomogeneous equation of state for dark energy fluid. The Hubble parameter described presents a periodic behavior such that early and late time acceleration are unified under the same mechanism. Also, it is considered a coupling between dark energy fluid, with homogeneous and constant EoS, and matter, that gives a periodic Universe too. The possible phantom phases and future singularities are studied in the oscillating Universe under discussion. The equivalent scalar-tensor representation for the same oscillating Universe is presented too.