Downloads & Free Reading Options - Results

The task of inferring a set of classes and class descriptions most likely to explain a given data set can be placed on a firm theoretical foundation using Bayesian statistics. Within this framework and using various mathematical and algorithmic approximations, the AutoClass system searches for the most probable classifications, automatically choosing the number of classes and complexity of class descriptions. A simpler version of AutoClass has been applied to many large real data sets, has discovered new independently-verified phenomena, and has been released as a robust software package. Recent extensions allow attributes to be selectively correlated within particular classes, and allow classes to inherit or share model parameters though a class hierarchy. We summarize the mathematical foundations of AutoClass.

The task of inferring a set of classes and class descriptions most likely to explain a given data set can be placed on a firm theoretical foundation using Bayesian statistics. Within this framework and using various mathematical and algorithmic approximations, the AutoClass system searches for the most probable classifications, automatically choosing the number of classes and complexity of class descriptions. A simpler version of AutoClass has been applied to many large real data sets, has discovered new independently-verified phenomena, and has been released as a robust software package. Recent extensions allow attributes to be selectively correlated within particular classes, and allow classes to inherit or share model parameters though a class hierarchy. We summarize the mathematical foundations of AutoClass.

In this paper, a new technique is used to an unsupervised learning image classification based on integration between neutrosophic sets and optimization linear programming.

We use classical invariant theory to solve the biholomorphic equivalence problem for two families of plane curve singularities previously considered in the literature. Our calculations motivate an intriguing conjecture that proposes a way of extracting a complete set of invariants of homogeneous plane curve singularities from their moduli algebras.

A computational theory for classification of natural biosonar targets is developed based on the properties of an example stimulus ensemble. An extensive set of echoes (84 800) from four different foliages was transcribed into a spike code using a parsimonious model (linear filtering, half-wave rectification, thresholding). The spike code is assumed to consist of time differences (interspike intervals) between threshold crossings. Among the elementary interspike intervals flanked by exceedances of adjacent thresholds, a few intervals triggered by disjoint half-cycles of the carrier oscillation stand out in terms of resolvability, visibility across resolution scales and a simple stochastic structure (uncorrelatedness). They are therefore argued to be a stochastic analogue to edges in vision. A three-dimensional feature vector representing these interspike intervals sustained a reliable target classification performance (0.06% classification error) in a sequential probability ratio test, which models sequential processing of echo trains by biological sonar systems. The dimensions of the representation are the first moments of duration and amplitude location of these interspike intervals as well as their number. All three quantities are readily reconciled with known principles of neural signal representation, since they correspond to the center of gravity of excitation on a neural map and the total amount of excitation.

Authors, Robert Karplus, Cella Stendler Lavatelli Illustrates the developmental theory of Jean Piaget using a study of the intellectual development of children and a description of the processes that characterize a child's thinking at successive stages of development. Shows children responding to tasks that highlight different mental operations essential to classification Summary Illustrates the developmental theory of Jean Piaget, using a study of the intellectual development of children and a description of the processes that characterize a child's thinking at succe./ia metages of development. Shows children responding to tasks that highlight different mental operations essential to classification.

In this short note, we are going to classify the cosmological evolutions in the recently proposed, nonrelativistic gravitational theory, the Ho\v{r}ava-Lifshitz (HL) theory. We consider the original HL theory (theory I), and the modified version obtained by an analytic continuation of parameters (theory II). We discuss the possiblilities of cosmological evolutions with arbitrary matter.

The classification of isoparametric hypersurfaces with four principal curvatures in the sphere interplays in a deep fashion with commutative algebra, whose abstract and comprehensive nature might obscure a differential geometer's insight into the classification problem that encompasses a wide spectrum of geometry and topology. In this paper, we make an effort to bridge the gap by walking through the important part of commutative algebra central to the classification of such hypersurfaces, such that all the essential ideal-theoretic ingredients are laid out in a way as much intuitive, motivating and geometric with rigor maintained as possible. We then explain how we developed the technical side of the entailed ideal theory, pertinent to isoparametric hypersurfaces with four principal curvatures, for the classification done in our papers~\cite{CCJ},~\cite{Ch1} and~\cite{Ch3}.

The classification of isoparametric hypersurfaces with four principal curvatures in the sphere interplays in a deep fashion with commutative algebra, whose abstract and comprehensive nature might obscure a differential geometer's insight into the classification problem that encompasses a wide spectrum of geometry and topology. In this paper, we make an effort to bridge the gap by walking through the important part of commutative algebra central to the classification of such hypersurfaces, such that all the essential ideal-theoretic ingredients are laid out in a way as much intuitive, motivating and geometric with rigor maintained as possible. We then explain how we developed the technical side of the entailed ideal theory, pertinent to isoparametric hypersurfaces with four principal curvatures, for the classification done in our papers~\cite{CCJ},~\cite{Ch1} and~\cite{Ch3}.

A close relationship is demonstrated between the catastrophe theory, Petrov's algebraic classification of the gravitational fields and the theory of phase transitions.

We give a complete classification of all static, spherically symmetric solutions of the SU(2) Einstein-Yang-Mills theory with a positive cosmological constant. Our classification proceeds in two steps. We first extend solutions of the radial field equations to their maximal interval of existence. In a second step we determine the Carter-Penrose diagrams of all 4-dimensional space-times constructible from such radial pieces. Based on numerical studies we sketch a complete phase space picture of all solutions with a regular origin.

We give a complete classification of all static, spherically symmetric solutions of the SU(2) Einstein-Yang-Mills theory with a positive cosmological constant. Our classification proceeds in two steps. We first extend solutions of the radial field equations to their maximal interval of existence. In a second step we determine the Carter-Penrose diagrams of all 4-dimensional space-times constructible from such radial pieces. Based on numerical studies we sketch a complete phase space picture of all solutions with a regular origin.

A dataset has been classified by some unknown classifier into two types of points. What were the most important factors in determining the classification outcome? In this work, we employ an axiomatic approach in order to uniquely characterize an influence measure: a function that, given a set of classified points, outputs a value for each feature corresponding to its influence in determining the classification outcome. We show that our influence measure takes on an intuitive form when the unknown classifier is linear. Finally, we employ our influence measure in order to analyze the effects of user profiling on Google's online display advertising.

Various scenarios of the initial inflation of the universe are distinguished by the choice of a scalar field {\em potential} $U(\phi)$ which simulates a {\it temporarily} non--vanishing {\em cosmological term}. Our new method, which involves a reparametrization in terms of the Hubble expansion parameter $H$, provides a classification of allowed inflationary potentials and of the stability of the critical points. It is broad enough to embody all known {\it exact} solutions involving one scalar field as special cases. Inflation corresponds to the evolution of critical points of some catastrophe manifold. The coalescence of its nondegenerate critical points with the creation of a degenerate critical point corresponds the reheating phase of the universe. This is illustrated by several examples.

Motivated by Tukey classification problems and building on work in \cite{Dobrinen/Todorcevic11}, we develop a new hierarchy of topological Ramsey spaces $\mathcal{R}_{\alpha}$, $\alpha

The purpose of this study was to investigate whether success in the Science - A Process Approach (SAPA) process of classification designed for primary grade children is contingent upon the children's developmental level as defined by Piaget's theory. The investigators sought to determine whether children who had reached the concrete operational stage of intellectual development are more likely to succeed on SAPA exercises requiring multi-classification ability than their primary classmates who are at the defined pre-operational level. Thirty children were assigned to either pre-operational or concrete operational on the basis of two diagnostic instruments and their mental ages were determined with another instrument. Matched pairs, one from each developmental state, were formed with each child having a mental age score within one standard error of one another. These matched pairs were assigned randomly to one of two teacher instructional groups and taught nine sequentially-arranged classification exercises as prescribed by SAPA. A two-factor analysis of variance with developmental level crossed with teachers was used to analyze the scores from the nine individually administered competency measures accompanying the SAPA classification exercises. The concrete operational group performed significantly better than the pre-operational group in three exercises requiring hierarchical classification and in four exercises requiring exhaustive sorting. (Author/MH)

This paper describes a successful approach to the task of concept learning for heuristic classification. This task differs from the usual concept learning task in three ways. First, classifications must be explained, not simply reported. Second, a program for this task must accommodate incomplete case descriptions. Third, the program must learn domain-specific knowledge for inferring case features needed for classification. Section 2 summarizes this learning and classification task. The traditional approach to concept learning and classification relies on generalizations. It requires a strong domain theory both to summarize training cases with concept descriptions and to classify new cases using these descriptions. Section 3 argues that this approach is ineffective for many domains. An alternative approach relies on exemplars. concepts are learned by retaining exemplars, and new cases are classified by matching them with exemplars. Our learning and classification program, Protos, uses the exemplar-based approach. Section 4 describes Protos's design and its appropriateness for weak-theory domains. To evaluate the design, Protos was applied to the task in clinical audiology of identifying a patient's hearing disorder from symptoms, test results, and history. An expert clinician instructed Protos with 200 cases - a level of training comparable to that received by student clinicians.

We classify elements of a cluster modular group into three types. We characterize them in terms of fixed point property of the action on the tropical compactifications associated with the corresponding cluster ensemble. The characterization gives an analogue of the Nielsen-Thurston classification theory on the mapping class group of a surface.

Classification theory of elementary classes deals with first order (elementary) classes of structures (i.e. fixing a set T of first order sentences, we investigate the class of models of T with the elementary submodel notion). It tries to find dividing lines, prove their consequences, prove "structure theorems, positive theorems" on those in the "low side" (in particular stable and superstable theories), and prove "non-structure, complexity theorems" on the "high side". It has started with categoricity and number of non-isomorphic models. It is probably recognized as the central part of model theory, however it will be even better to have such (non-trivial) theory for non-elementary classes. Note also that many classes of structures considered in algebra are not first order; some families of such classes are close to first order (say have kind of compactness). But here we shall deal with a classification theory for the more general case without assuming knowledge of the first order case (and in most parts not assuming knowledge of model theory at all). The present paper includes an introduction to the forthcoming book on Classification Theory for Abstract Elementary Classes

lambda-good frame is for us a parallel of the class of models of a superstable theory. Our main line is to start with lambda-good^+ frame s, categorical in lambda, n-successful for n large enough and try to have parallel of stability theory for K_{s(+l)} for l=lambda. Recall there are reasonable lambda-frames which are not n-excellent but still we can say alot on models in $K_{s(+l)} for l

lambda-good frame is for us a parallel of the class of models of a superstable theory. Our main line is to start with lambda-good^+ frame s, categorical in lambda, n-successful for n large enough and try to have parallel of stability theory for K_{s(+l)} for l=lambda. Recall there are reasonable lambda-frames which are not n-excellent but still we can say alot on models in $K_{s(+l)} for l

We study ramified covers of the projective plane. Given a smooth projective surface S and a generic enough projection of S to the projective plane, we get a cover of the plane ramified over a plane curve. The branch curve is usually singular, but is classically known to have only cusps and nodes as singularities for a generic projection. Several questions arise. First, what is the_geography_ of branch curves among all nodal-cuspidal curves? Second: what is the_geometry_ of branch curves? In other words, how can one distinguish a branch curve from a non-branch curve with the same numerical invariants? For example, a plane sextic curve with six cusps is known to be a branch curve of a generic projection iff its six cusps lie on a conic curve, i.e., form a special 0-cycle on the plane. We start with reviewing what is known about the answers to these two questions, mentioning both simple and some non-trivial results. We continue with study of classical work of Beniamino Segre which gives a complete answer to the second question in the case when S is a smooth surface in a three-dimensional projective space. We give an interpretation of Segre's work in terms of a study of Picard group of 0-cycles on a singular plane curve B. We also review examples of small degree. The Appendix written by Eugenii Shustin shows the existence of many new Zariski pairs of plane curves. We hope to continue this paper with a generalization to the case of smooth surfaces in a projective space of any dimension.

The well-known difficulties arising in a classification which is not set-theoretically trivial---involving what is sometimes called a non-smooth quotient---have been overcome in a striking way in the theory of operator algebras by the use of what might be called a classification functor---the very existence of which is already a surprise. Here the notion of such a functor is developed abstractly, and a number of examples are considered (including those which have arisen for various classes of operator algebras).

Classification and construction of symmetry protected topological (SPT) phases in interacting boson and fermion systems have become a fascinating theoretical direction in recent years. It has been shown that the (generalized) group cohomology theory or cobordism theory can give rise to a complete classification of SPT phases in interacting boson/spin systems. Nevertheless, the construction and classification of SPT phases in interacting fermion systems are much more complicated, especially in 3D. In this work, we revisit this problem based on the equivalent class of fermionic symmetric local unitary (FSLU) transformations. We construct very general fixed point SPT wavefunctions for interacting fermion systems. We naturally reproduce the partial classifications given by special group super-cohomology theory, and we show that with an additional $\tilde{B}H^2(G_b, \mathbb Z_2)$ (the so-called obstruction free subgroup of $H^2(G_b, \mathbb Z_2)$) structure, a complete classification of SPT phases for three-dimensional interacting fermion systems with a total symmetry group $G_f=G_b\times \mathbb Z_2^f$ can be obtained for unitary symmetry group $G_b$. We also discuss the procedure of deriving a general group super-cohomology theory in arbitrary dimensions.

In this paper we constructed new model of plastic deformation. The knot theory was used to classify the plastic state.

As Lie algebras of compact connected Lie groups, semisimple Lie algebras have wide applications in the description of continuous symmetries of physical systems. Mathematically, semisimple Lie algebra admits a Cartan-Weyl basis of generators which consists of a Cartan subalgebra of mutually commuting generators H_I and a number of step generators E^\alpha that are characterized by a root space of non-degenerate one-forms \alpha. This simple decomposition in terms of the root space allows for a complete classification of semisimple Lie algebras. In this paper, we introduce the analogous concept of a Cartan-Weyl Lie 3-algebra. We analyze their structure and obtain a complete classification of them. Many known examples of metric Lie 3-algebras (e.g. the Lorentzian 3-algebras) are special cases of the Cartan-Weyl 3-algebras. Due to their elegant and simple structure, we speculate that Cartan-Weyl 3-algebras may be useful for describing some kinds of generalized symmetries. As an application, we consider their use in the Bagger-Lambert-Gustavsson (BLG) theory.

We investigate BPS conditions preserving n/12 (n = 1,...,6) supersymmetries in the Aharony-Bergman-Jafferis-Maldacena (ABJM) model. The BPS equations are classified in terms of the number of preserved supercharges and remaining subgroups of the SU(4)_R symmetry. We study structures of a map between projection conditions for the supercharges in eleven dimensions and those in the ABJM model. The BPS configurations in the ABJM model can be interpreted as known BPS objects in eleven-dimensional Mtheory,such as intersecting M2, M5-branes, M-waves, KK-monopoles and M9-branes. We also show that these BPS conditions reduce to those in N = 8 super Yang-Mills theory via the standard D2-reduction procedure in a consistent way with the M-theory interpretation of the BPS conditions.

We show that a number of results on abstract elementary classes (AECs) hold in accessible categories with concrete directed colimits. In particular, we prove a generalization of a recent result of Boney on tameness under a large cardinal assumption. We also show that such categories support a robust version of the Ehrenfeucht-Mostowski construction. This analysis has the added benefit of producing a purely language-free characterization of AECs, and highlights the precise role played by the coherence axiom.

This is the first of a series of papers in which we systematically use singularity theory to study four dimensional N=2 superconformal field theories. Our main focus in this paper is to identify what kind of singularity is needed to define a SCFT. The constraint for a hypersurface singularity has been found by Sharpere and Vafa, and here the complete set of solutions are listed using a related mathematical result of Stephen S. T. Yau and Yu. We also study other type of singularities such as the complete intersection, quotient of hypersurface singularity by a finite group and non-isolated singularity. We finally conjecture that any three dimensional rational Gorenstein graded isolated singularity should define a N=2 SCFT. We explain how to extract various interesting physical quantities such as Seiberg-Witten geometry, central charges, exact marginal deformations, BPS quiver, RG flow trajectory, etc from the properties of singularity.

This is the first of a series of papers in which we systematically use singularity theory to study four dimensional N=2 superconformal field theories. Our main focus in this paper is to identify what kind of singularity is needed to define a SCFT. The constraint for a hypersurface singularity has been found by Sharpere and Vafa, and here the complete set of solutions are listed using a related mathematical result of Stephen S. T. Yau and Yu. We also study other type of singularities such as the complete intersection, quotient of hypersurface singularity by a finite group and non-isolated singularity. We finally conjecture that any three dimensional rational Gorenstein graded isolated singularity should define a N=2 SCFT. We explain how to extract various interesting physical quantities such as Seiberg-Witten geometry, central charges, exact marginal deformations, BPS quiver, RG flow trajectory, etc from the properties of singularity.

Relying upon the division-algebra classification of Clifford algebras and spinors, a classification of generalized supersymmetries (or, with a slight abuse of language,"generalized supertranslations") is provided. In each given space-time the maximal, saturated, generalized supersymmetry, compatible with the division-algebra constraint that can be consistently imposed on spinors and on superalgebra generators, is furnished. Constraining the superalgebra generators in both the complex and the quaternionic cases gives rise to the two classes of constrained hermitian and holomorphic generalized supersymmetries. In the complex case these two classes of generalized supersymmetries can be regarded as complementary. The quaternionic holomorphic supersymmetry only exists in certain space-time dimensions and can admit at most a single bosonic scalar central charge. The results here presented pave the way for a better understanding of the various $M$ algebra-type of structures which can be introduced in different space-time signatures and in association with different division algebras, as well as their mutual relations. In a previous work, e.g., the introduction of a complex holomorphic generalized supersymmetry was shown to be necessary in order to perform the analytic continuation of the standard $M$-theory to the 11-dimensional Euclidean space. As an application of the present results, it is shown that the above algebra also admits a 12-dimensional, Euclidean, $F$-algebra presentation.

In 2005 V. Turaev introduced the theory of topology of words and phrases. Turaev defined an equivalence relation on generalized words and phrases which is called homotopy. This is suggested by the Reidemeister moves in the knot theory. Then Turaev gave the homotopy classification of generalized words with less than or equal to five letters. In this paper we give the classification of generalized phrases up to homotopy with less than or equal to three letters. To do this we construct a new homotopy invariant for nanophrases over any $\alpha$.

In 2005 V. Turaev introduced the theory of topology of words and phrases. Turaev defined an equivalence relation on generalized words and phrases which is called homotopy. This is suggested by the Reidemeister moves in the knot theory. Then Turaev gave the homotopy classification of generalized words with less than or equal to five letters. In this paper we give the classification of generalized phrases up to homotopy with less than or equal to three letters. To do this we construct a new homotopy invariant for nanophrases over any $\alpha$.

http://uf.catalog.fcla.edu/uf.jsp?st=UF002641407&ix=pm&I=0&V=D&pm=1

http://uf.catalog.fcla.edu/uf.jsp?st=UF002641407&ix=pm&I=0&V=D&pm=1

http://uf.catalog.fcla.edu/uf.jsp?st=UF002641407&ix=pm&I=0&V=D&pm=1

A general theory for the tracking and classification of multiple targets using a Bayesian approach is presented, together with its specialization to independent, identically distributed target models. The implementation of the theory is through the Generalized Tracker/Classifier. Simulation results to illustrate the algorithm are also given.

Let $\lambda=(\lambda_1,\lambda_2,...)$ be a \emph{partition} of $n$, a sequence of positive integers in non-increasing order with sum $n$. Let $\Omega:=\{1,...,n\}$. An ordered partition $P=(A_1,A_2,...)$ of $\Omega$ has \emph{type} $\lambda$ if $|A_i|=\lambda_i$. Following Martin and Sagan, we say that $G$ is \emph{$\lambda$-transitive} if, for any two ordered partitions $P=(A_1,A_2,...)$ and $Q=(B_1,B_2,...)$ of $\Omega$ of type $\lambda$, there exists $g\in G$ with $A_ig=B_i$ for all $i$. A group $G$ is said to be \emph{$\lambda$-homogeneous} if, given two ordered partitions $P$ and $Q$ as above, inducing the sets $P'=\{A_1,A_2,...\}$ and $Q'=\{B_1,B_2,...\}$, there exists $g\in G$ such that $P'g=Q'$. Clearly a $\lambda$-transitive group is $\lambda$-homogeneous. The first goal of this paper is to classify the $\lambda$-homogeneous groups. The second goal is to apply this classification to a problem in semigroup theory. Let $\trans$ and $\sym$ denote the transformation monoid and the symmetric group on $\Omega$, respectively. Fix a group $H\leq \sym$. Given a non-invertible transformation $a\in \trans\setminus \sym$ and a group $G\leq \sym$, we say that $(a,G)$ is an \emph{$H$-pair} if the semigroups generated by $\{a\}\cup H$ and $\{a\}\cup G$ contain the same non-units, that is, $ < a,G>\setminus G= < a,H>\setminus H$. Using the classification of the $\lambda$-homogeneous groups we classify all the $\sym$-pairs. This topic involves both group theory and semigroup theory; we have attempted to include enough exposition to make the paper self-contained for researchers in both areas. The paper finishes with a number of open problems on permutation and linear groups.

The long range goal of this project was the development of new signal processing and data classification methods based on ideas from the field of nonlinear dynamics. Specifically, nonlinear time-domain modeling methods using closed-form dynamical equations were to be formulated in a rigorous way in the framework of standard statistical signal classification methods. The aim was to provide new algorithms and automated tools which could enhance existing Navy systems for Sonar detection and classification, and to eventually transition them to the fleet.

The long range goal of this project was the development of new signal processing and data classification methods based on ideas from the field of nonlinear dynamics. Specifically, nonlinear time-domain modeling methods using closed-form dynamical equations were to be formulated in a rigorous way in the framework of standard statistical signal classification methods. The aim was to provide new algorithms and automated tools which could enhance existing Navy systems for Sonar detection and classification, and to eventually transition them to the fleet.

We propose a quantum field theory (QFT) method to approach the classification of indefinite sector of Kac-Moody algebras. In this approach, Vinberg relations are interpreted as the discrete version of the QFT_{2} equation of motion of a scalar field and Dynkin diagrams as QFT_{2} Feynman graphs. In particular, we show that Dynkin diagrams of su(n+1) series (n\geq 1) can be interpreted as free field propagators and T_{p,q,r} diagrams as the vertex of \phi^{3} interaction. Other results are also given.

We propose a quantum field theory (QFT) method to approach the classification of indefinite sector of Kac-Moody algebras. In this approach, Vinberg relations are interpreted as the discrete version of the QFT_{2} equation of motion of a scalar field and Dynkin diagrams as QFT_{2} Feynman graphs. In particular, we show that Dynkin diagrams of su(n+1) series (n\geq 1) can be interpreted as free field propagators and T_{p,q,r} diagrams as the vertex of \phi^{3} interaction. Other results are also given.

We propose a quantum field theory (QFT) method to approach the classification of indefinite sector of Kac-Moody algebras. In this approach, Vinberg relations are interpreted as the discrete version of the QFT_{2} equation of motion of a scalar field and Dynkin diagrams as QFT_{2} Feynman graphs. In particular, we show that Dynkin diagrams of su(n+1) series (n\geq 1) can be interpreted as free field propagators and T_{p,q,r} diagrams as the vertex of \phi^{3} interaction. Other results are also given.

We propose a quantum field theory (QFT) method to approach the classification of indefinite sector of Kac-Moody algebras. In this approach, Vinberg relations are interpreted as the discrete version of the QFT_{2} equation of motion of a scalar field and Dynkin diagrams as QFT_{2} Feynman graphs. In particular, we show that Dynkin diagrams of su(n+1) series (n\geq 1) can be interpreted as free field propagators and T_{p,q,r} diagrams as the vertex of \phi^{3} interaction. Other results are also given.

AFB Research series : no. 25

Human beings judge patterns (such as images) by complex mental processes, some of which may not be known, while computing machines extract features. By representing the human judgements with simple measurements and reducing them and the machine extracted features to a common metric space and fitting them by regression, the judgements of human experts rendered on a sample of patterns may be imposed on a pattern population to provide automatic classification.

Human beings judge patterns (such as images) by complex mental processes, some of which may not be known, while computing machines extract features. By representing the human judgements with simple measurements and reducing them and the machine extracted features to a common metric space and fitting them by regression, the judgements of human experts rendered on a sample of patterns may be imposed on a pattern population to provide automatic classification.

We study thermodynamics of cosmological models in the Horava-Lifshitz theory of gravity, and systematically investigate the evolution of the universe filled with a perfect fluid that has the equation of state $p=w\rho$, where $p$ and $\rho$ denote, respectively, the pressure and energy density of the fluid, and $w$ is an arbitrary real constant. Depending on specific values of the free parameters involved in the models, we classify all of them into various cases. In each case the main properties of the evolution are studied in detail, including the periods of deceleration and/or acceleration, and the existence of big bang, big crunch, and big rip singularities. We pay particular attention on models that may give rise to a bouncing universe.

We study thermodynamics of cosmological models in the Horava-Lifshitz theory of gravity, and systematically investigate the evolution of the universe filled with a perfect fluid that has the equation of state $p=w\rho$, where $p$ and $\rho$ denote, respectively, the pressure and energy density of the fluid, and $w$ is an arbitrary real constant. Depending on specific values of the free parameters involved in the models, we classify all of them into various cases. In each case the main properties of the evolution are studied in detail, including the periods of deceleration and/or acceleration, and the existence of big bang, big crunch, and big rip singularities. We pay particular attention on models that may give rise to a bouncing universe.