arXiv subjects

Jan Žemlička

Publications and source records attributed to Jan Žemlička.

12 recordsLinked to original sources

arXiv2026

Hereditarily and Super Bassian Modules over Certain Rings

We characterize in certain basic cases when a module over a ring is either {\it hereditarily Bassian} or {\it super Bassian} in the sense that either each its proper submodule is Bassian or, respectively, each its proper epimorphic image is Bassian. We prove several structural criteria for both hereditarily Bassian and super Bassian modules over non-primitive Dedekind prime rings, and in particular Dedekind domains. Over these rings, we establish that a singular module is super Bassian exactly when it is Bassian, which is true if and only if it is Bassian. In addition, for an arbitrary (not necessarily singular) module over a non-primitive Dedekind prime ring, the property of being super Bassian curiously implies the property of being hereditary Bassian always. Our results somewhat continue and supply recent results due to Tuganbaev in Mathematics (2026) and Blacher in J. Algebra (2026).

math.RA↗

arXiv2026

Deep Learning in the Sequence Space

We develop a deep learning algorithm for constructing globally accurate approximations to functional rational expectations equilibria of dynamic stochastic economies in the sequence space. We use deep neural networks to parameterize key equilibrium objects, such as policies or prices, as functions of truncated histories of exogenous shocks. We train the neural networks to satisfy equilibrium conditions along simulated paths of the economy. We illustrate the performance of our method in three environments: (i) a high-dimensional overlapping generations economy with multiple sources of aggregate risk; (ii) an economy with heterogeneous households and firms facing uninsurable idiosyncratic risk and large shocks to idiosyncratic and aggregate volatility; and (iii) a stochastic life-cycle economy with a continuous asset choice and a discrete early-retirement choice that induces local convexities in the continuation values of working-age cohorts. We also propose practical neural policy architectures that guarantee monotonicity of predicted policies, enabling the endogenous grid method to simplify parts of the algorithm. We achieve high precision throughout, with the mean error in equilibrium conditions below $0.2\%$.

econ.GN↗

arXiv2023

Economics-Inspired Neural Networks with Stabilizing Homotopies

Contemporary deep learning based solution methods used to compute approximate equilibria of high-dimensional dynamic stochastic economic models are often faced with two pain points. The first problem is that the loss function typically encodes a diverse set of equilibrium conditions, such as market clearing and households' or firms' optimality conditions. Hence the training algorithm trades off errors between those -- potentially very different -- equilibrium conditions. This renders the interpretation of the remaining errors challenging. The second problem is that portfolio choice in models with multiple assets is only pinned down for low errors in the corresponding equilibrium conditions. In the beginning of training, this can lead to fluctuating policies for different assets, which hampers the training process. To alleviate these issues, we propose two complementary innovations. First, we introduce Market Clearing Layers, a neural network architecture that automatically enforces all the market clearing conditions and borrowing constraints in the economy. Encoding economic constraints into the neural network architecture reduces the number of terms in the loss function and enhances the interpretability of the remaining equilibrium errors. Furthermore, we present a homotopy algorithm for solving portfolio choice problems with multiple assets, which ameliorates numerical instabilities arising in the context of deep learning. To illustrate our method we solve an overlapping generations model with two permanent risk aversion types, three distinct assets, and aggregate shocks.

econ.GN↗

arXiv2022

Connected objects in categories of $S$-acts

In this paper, the categorial property of compactness of an object, i. e. commuting of the corresponding $\Hom$ functor with coproducts, is studied in categories of $S$-acts and the corresponding structural properties of compact $S$-acts are shown. In order to establish a general context and to unify the approach to both of the most important categories of $S$-acts, the notion of a concrete category with unique decomposition of objects is introduced and studied.

math.CT↗

arXiv2021

Perfect monoids with zero and categories of $S$-acts

In this paper, we study the relationship between the two main categories of $S$-acts for a monoid $S$ with zero from the viewpoint of existence of projective covers and the equivalence is proven. Furthermore, monoids with zeros over which all compact acts are cyclic are characterized.

math.GR↗

arXiv2021

Self-small products of abelian groups

For abelian groups $A, B$, $A$ is called $B$-small if the covariant functor $Hom(A,-)$ commutes with all direct sums $B^{(κ)}$ and $A$ is self-small provided it is $A$-small. The paper characterizes self-small products applying developed closure properties of the classes of relatively small groups. As a consequence, self-small products of finitely generated abelian groups are described.

math.GR↗

arXiv2021

Autocompact objects of Ab5 categories

The aim of the paper is to describe autocompact objects in Ab5-categories, i.e. objects in cocomplete abelian categories with exactness preserving filtered colimits of exact sequences, whose covariant Hom-functor commutes with copowers of the object itself. A characterization of non-autocompact object is given, a general criterion of autocompactness of an object via the structure of its endomorphism ring is presented and a criterion of autocompactness of products is proven.

math.CT↗

arXiv2020

Self-injective von Neumann regular rings and Köthe's Conjecture

One of the many equivalent formulation of the Köthe's conjecture is the assertion that there exists no ring which contains two nil right ideals whose sum is not nil. We discuss several consequences of an observation that if the Koethe conjecture fails then there exists a counterexample in the form of a countable local subring of a suitable self-injective prime (von Neumann) regular ring.

math.RA↗

arXiv2017

Compactness in abelian categories

We relativize the notion of a compact object in an abelian category with respect to a fixed subclass of objects. We show that the standard closure properties persist to hold in this case. Furthermore, we describe categorical and set-theoretical conditions under which all products of compact objects remain compact.

math.CT↗

arXiv2017

Torsion classes generated by silting modules

We study the classes of modules which are generated by a silting module. In the case of either hereditary or perfect rings it is proved that these are exactly the torsion $\mathcal{T}$ such that the regular module has a special $\mathcal{T}$-preenvelope. In particular every torsion enveloping class in $\textrm{Mod-} R$ are of the form $\mathrm{Gen}(T)$ for a minimal silting module $T$. For the dual case we obtain for general rings that the covering torsion-free classes of modules are exactly the classes of the form $\mathrm{Cogen}(T)$, where $T$ is a cosilting module.

math.RT↗

arXiv2014

The defect functor of a homomorphism and direct unions

We will study commuting properties of the defect functor $\mathrm{Dev}_β=\mathrm{Coker}\mathrm{Hom}_\mathcal{C}(β,-)$ associate to a homomorphism $β$ in a finitely presented category. As an application, we characterize objects $M$ such that $\mathrm{Ext}^1_\mathcal{C}(M,-)$ commutes with direct unions (i.e. direct limits of monomorphisms), assuming that $\mathcal{C}$ has a generator which is a direct sum of finitely presented projective objects.

math.RT↗

arXiv2014

Algebraic properties of word equations

The question about maximal size of independent system of word equations is one of the most striking problems in combinatorics on words. Recently, Aleksi Saarela has introduced a new approach to the problem that is based on linear-algebraic properties of polynomials encoding the equations and their solutions. In this paper we develop further this approach and take into account other algebraic properties of polynomials, namely their factorization. This, in particular, allows to improve the bound for the number of independent equations with maximal rank from quadratic to linear.

cs.FL↗