Browsing by Title

Sort by: Order: Results:

Now showing items 1-20 of 23
  • Saukonoja, Teemu; Pasanen, Tomi (2008)
    Artificial intelligence research in Texas Hold’em poker has recently mainly focused on heads-up fixed limit game. Game theoretic methods used in the poker agents capable of playing at the very best level are not easily generalized to other forms of Texas Hold’em poker. In this paper we present a general approach to build a poker agent, where the betting strategy is defined by estimating the expected values of the actions. This approach suites well larger number of players and it is also easily modified to suite no limit games. We focus on defining of the evaluation function. It has a very crucial part at the overall level of the poker agent in this kind of solution.
  • Norta, Alex (2011)
    The management and coordination of business-process collaboration experiences changes because of globalization, specialization, and innovation. Service-oriented computing (SOC) is a means towards businessprocess automation and recently, many industry standards emerged to become part of the service-oriented architecture (SOA) stack. In a globalized world, organizations face new challenges for setting up and carrying out collaborations in semi-automating ecosystems for business services. For being efficient and effective, many companies express their services electronically in what we term business-process as a service (BPaaS). Companies then source BPaaS on the fly from third parties if they are not able to create all service-value inhouse because of reasons such as lack of reasoures, lack of know-how, cost- and time-reduction needs. Thus, a need emerges for BPaaS-HUBs that not only store service offers and requests together with information about their issuing organizations and assigned owners, but that also allow an evaluation of trust and reputation in an anonymized electronic service marketplace. In this paper, we analyze the requirements, design architecture and system behavior of such a BPaaS-HUB to enable a fast setup and enactment of business-process collaboration. Moving into a cloud-computing setting, the results of this paper allow system designers to quickly evaluate which services they need for instantiationg the BPaaS-HUB architecture. Furthermore, the results also show what the protocol of a backbone service bus is that allows a communication between services that implement the BPaaS-HUB. Finally, the paper analyzes where an instantiation must assign additional computing resources vor the avoidance of performance bottlenecks.
  • Yu, Huizhen; Rousu, Juho (2008)
    We consider structured prediction problems with a parametrized linear prediction function, and the associated parameter optimization problems in large margin type of discriminative training. We propose a dual optimization approach which uses the restricted simplicial decomposition method to optimize a reparametrized dual problem. Our reparametrization reduces the dimension of the space of the dual function to one that is linear in the number of parameters and training examples, and hence independent of the dimensionality of the prediction outputs. This in conjunction with simplicial decomposition makes our approach efficient. We discuss the connections of our approach with related earlier works, and we show its advantages.
  • Yu, Huizhen (2010)
    We consider approximate policy evaluation for finite state and action Markov decision processes (MDP) in the off-policy learning context and with the simulation-based least squares temporal difference algorithm, LSTD(λ). We establish for the discounted cost criterion that the off-policy LSTD(λ) converges almost surely under mild, minimal conditions. We also analyze other convergence and boundedness properties of the iterates involved in the algorithm, and based on them, we suggest a modification in its practical implementation. Our analysis uses theories of both finite space Markov chains and Markov chains on topological spaces, in particular, the e-chains.
  • Järvinen, Ilpo; Chemmagate, Binoy; Ding, Aaron Yi; Daniel, Laila; Kojo, Markku (University of Helsinki, Department of Computer Science, 2012)
    This experimental study analyzes the effects of larger TCP initial window on competing interactive media and Web traffic in a larger number of cellular access configurations. In addition, we analyze the effect of shorter initial RTO on TCP performance in cellular access configurations. Both simulation and real network experiments were conducted. The initial window of ten segments reduces TCP elapsed times when the number of flows is small enough, however, with large number of flows it introduces losses that require TCP timeout. The initial RTO change from three to one second improves elapsed time in limited number of configurations, but in other cellular configurations spurious timeouts trigger almost alway during TCP three-way handshake due to the lower timeout.
  • Hinkkanen, Tero; Kurhila, Jaakko; Pasanen, Tomi A. (2008)
    We present a framework for evaluating believability of characters in first-person shooter (FPS) games and look into the development of non-player character’s user-perceived believability. The used framework is composed of two aspects: firstly, character movement and animation, secondly, behavior. Examination of three different FPS games yields that the newer the game was, the better the believability of characters in the game. Moreover, the results from both the aspects of the framework were mutually balanced through all games examined.
  • Huizhen, Yu (2010)
    We consider approximate policy evaluation for finite state and action Markov decision processes (MDP) with the least squares temporal difference algorithm, LSTD(λ), in an explorationenhanced off-policy learning context. We establish for the discounted cost criterion that the off-policy LSTD(λ) converges almost surely under mild, minimal conditions. We also analyze other convergence and boundedness properties of the iterates involved in the algorithm. Our analysis draws on theories of both finite space Markov chains and weak Feller Markov chains on topological spaces. Our results can be applied to other temporal difference algorithms and MDP models. As examples, we give a convergence analysis of an off-policy TD(λ) algorithm and extensions to MDP with compact action and state spaces.
  • Bingham, Ella; Koivisto, Mikko; Leino, Yrjö; Mannila, Heikki (2010)
    Linkage disequilibrium (LD) refers to the statistical dependency of the DNA content at nearby locations of the chromosome. Numerous approaches to analyze genome data rely on the well documented fact that LD decays monotonously with the distance of the studied loci. This decay, though noisy and modi ed by a number of factors, can be attributed to the recombination process, a major source of genetic variation in diploid organisms. In this work we take first steps toward analyzing the extent of LD between very distant loci, even loci from different chromosomes. This is in contrast to traditional "genome-wide" analyses which merely study the LD within each chromosome separately. We design several measures of LD, and use them for analyzing the HapMap data. We also consider LD between supermarkers determined by haplotype clusters in windows of a few SNPs. We report on suggestive pairs of loci where unusually large correlations are observed within all ethnic groups. We describe how the computations can be arranged in a way that enables an all-pairs analysis of the data, that is, all pairs of loci across all the 22 autosomal chromosomes. This kind of "genome times genome" analysis is computationally very burdensome due to the sheer number of possible pairs. We show ways to make it feasible.
  • Yu, Huizhen; Bertsekas, Dimitri P. (2008)
    We consider linear fixed point equations and their approximations by projection on a low dimensional subspace. We derive new bounds on the approximation error of the solution, which are expressed in terms of low dimensional matrices and can be computed by simulation. When the fixed point mapping is a contraction, as is typically the case in Markov decision processes (MDP), one of our bounds is always sharper than the standard contraction-based bounds, and another one is often sharper. The former bound is also tight in a worst-case sense. Our bounds also apply to the non-contraction case, including policy evaluation in MDP with nonstandard projections that enhance exploration. There are no error bounds currently available for this case to our knowledge.
  • Suomalainen, Lauri; Nikkhouy, Emad; Ding, Aaron Yi; Tarkoma, Sasu (University of Helsinki, Department of Computer Science, 2014)
    Software-Defined Networking (SDN) is a novel solution to network configuration and management. Its openness and programmability features have greatly motivated the open source communities where numerous applications and tools are developed for various R&D purposes. For the strength of SDN, the upcoming 5th Generation mobile networks (5G) can also benefit from the modular and open design to innovate the network architecture and services. In this report, we present a survey of existing open source platforms, applications and tools for SDN and 5G research. We discuss the potential directions and share our perspectives in this domain.
  • Galbrun, Esther (2010)
    Phrase-Based Statistical Machine Translation systems model the translation process using pairs of corresponding sequences of words extracted from parallel corpora. These biphrases are stored in phrase tables that typically contain several millions such entries, making it di cult to assess their quality without going to the end of the translation process. Our work is based on the examplifying study of phrase tables generated from the Europarl data, from French to English. We give some statistical information about the biphrases contained in the phrase table, evaluate the coverage of previously unseen sentences and analyse the e ects of pruning on the translation.
  • Rissanen, Jorma; Myllymäki, Petri; Roos, Teemu; Santhanam, Narayana Prasad (University of Helsinki, Department of Computer Science, 2014)
  • Fagerholm, Fabian; Paasivaara, Maria; Jedlitschka, Andreas; Kuvaja, Pasi; Kuhrmann, Marco; Männistö, Tomi; Münch, Jürgen; Raatikainen, Mikko (University of Helsinki, Department of Computer Science, 2014)
  • Bertsekas, Dimitri P.; Yu, Huizhen (2010)
    We consider the classical nite-state discounted Markovian decision problem, and we introduce a new policy iteration-like algorithm for fi nding the optimal Q-factors. Instead of policy evaluation by solving a linear system of equations, our algorithm requires (possibly inexact) solution of a nonlinear system of equations, involving estimates of state costs as well as Q-factors. This is Bellman's equation for an optimal stopping problem that can be solved with simple Q-learning iterations, in the case where a lookup table representation is used; it can also be solved with the Q-learning algorithm of Tsitsiklis and Van Roy [TsV99], in the case where feature-based Q-factor approximations are used. In exact/lookup table representation form, our algorithm admits asynchronous and stochastic iterative implementations, in the spirit of asynchronous/modi ed policy iteration, with lower overhead and more reliable convergence advantages over existing Q-learning schemes. Furthermore, for large-scale problems, where linear basis function approximations and simulation-based temporal di erence implementations are used, our algorithm resolves e ffectively the inherent difficulties of existing schemes due to inadequate exploration.
  • Bertsekas, Dimitri P.; Huizhen, Yu (2010)
    We consider the classical finite-state discounted Markovian decision problem, and we introduce a new policy iteration-like algorithm for finding the optimal Q-factors. Instead of policy evaluation by solving a linear system of equations, our algorithm requires (possibly inexact) solution of a nonlinear system of equations, involving estimates of state costs as well as Q-factors. This is Bellman's equation for an optimal stopping problem that can be solved with simple Q-learning iterations, in the case where a lookup table representation is used; it can also be solved with the Q-learning algorithm of Tsitsiklis and Van Roy [TsV99], in the case where feature-based Q-factor approximations are used. In exact/lookup table representation form, our algorithm admits asynchronous and stochastic iterative implementations, in the spirit of asynchronous/modified policy iteration, with lower overhead and/or more reliable convergence advantages over existing Q-learning schemes. Furthermore, for large-scale problems, where linear basis function approximations and simulation-based temporal difference implementations are used, our algorithm resolves effectively the inherent difficulties of existing schemes due to inadequate exploration.
  • Pasanen, Tomi (2008)
    We consider random binary search trees when the input consists of a multiset, i.e. a set with multiple occurrences of equal elements, and prove that the randomized insertion and deletion algorithms produce random search trees regardless of multiplicities; even if all the elements are equal during the tree updates, a search tree will maintain its randomness. Thus, equal elements do not degenerate a random search tree and they need not to be handled in any special way. We consider also stability of a search tree with respect to its inorder traversal and prove that the algorithms used produce stable trees. This introduces an implicit indexing of equal elements giving another proof that multiplicities do not pose problems when maintaining random binary search trees.
  • Mäkinen, Veli; Navarro, Gonzalo; Sirén, Jouni; Välimäki, Niko (2008)
    A repetitive sequence collection is one where portions of a base sequence of length n are repeated many times with small variations, forming a collection of total length N. Examples of such collections are version control data and genome sequences of individuals, where the differences can be expressed by lists of basic edit operations. Flexible and efficient data analysis on a such typically huge collection is plausible using suffix trees. However, suffix tree occupies O(N logN) bits, which very soon inhibits in-memory analyses. Recent advances in full-text indexing reduce the space of suffix tree to NHk + o(N log σ) bits at the cost of running times of its operations increasing by polylog(N) factor. Here Hk is the k-th order entropy of the collection and σ is the alphabet size. Notice that for r identical copies of an incompressible base sequence, the bound simplifies to N log σ(1 + o(1)) bits. We develop new static/dynamic full-text self-indexes based on the run-length encoding whose space-requirements are much less dependent on N. For example, we obtain an index occupying Rlog σ(1+o(1))+R log N R (1+o(1))+r log n+O((s+r) log(s+r)) bits, where s is the total number of basic edit operations to convert the r repeats into substrings of the base sequence, and R ≤ min(n, nHk)+O((s+r) logσ N), where the O() term holds in the expected case. The new indexes can be plugged into a recent dynamic fully-compressed suffix tree using an additional O((N/δ) log N) bits of space for any δ = polylog(N), and retaining the polylog(N) time slowdown on operations.
  • Mäkinen, Veli; Sirén, Jouni; Välimäki, Niko (2008)
    In the near future, biomolecular engineering techniques will reach a state where the sequencing of individual genomes becomes feasible. This progress will create huge expectations for the data analysis domain to reveal new knowledge on the ”secrets of life”. Quite rudimentary reasons may inhibit such breakthroughs; it may not be feasible to store all the data in a form that would enable anything but most basic data analysis routines to be executed. This paper is devoted into studying ways to store massive sets of complete individual genomes in space-efficient manner so that retrieval of the content as well as queries on the content of the sequences can be provided time-efficiently. We show that although the state-of-the-art full-text self-indexes do not yet provide satisfactory space bounds for this specific task, after carefully engineering those structures it is possible to achieve very attractive results; the new structures are fully able to exploit the fact that the individual genomes are highly similar. We confirm the theoretical findings by experiments on large DNA sequences, and also on version control data, that forms another application domain for our methods.