In this chapter we present how the Whittle's indexation approach, originally developed for the restless bandits problem, can be used to address multi-objective stochastic optimization problems. For clarity, we focus on Markov decision processes with two objectives, two states and two actions, and provide an optimal solution in terms of Whittle indices. We then broadly discuss the applications of our solution in optimal scheduling of queueing systems with abandonments and optimal choice of venture capitalists investments. We believe there is a need to develop the Whittle index theory further in order to tackle more general multi-objective problems.
To read the full-text of this research,
Request full-text PDF
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.
January 2002 · Lecture Notes in Computer Science
In this paper, we address a data-mining problem that is the discovery of local sequential patterns from a set of long sequences.
Each local sequential pattern is represented by a pattern A→B and a time period in which A→B is frequent. Such patterns are actually very common in practice and are potentially very useful. However it is impractical
to use traditional methods on this problem directly.
... [Show full abstract] We propose a suffix-tree-like data structure for indexing the instances
of the patterns. Based on this index, our mining method can discover all locally frequent patterns after one scan of the sequences.
We have analyzed the behavior of the problem and evaluated the performance of our algorithm with both synthetic and real data.
The results correspond with the definition of the problem and verify the superiority of our approach. Read more February 1990 · Mathematics of Operations Research
The authors' work on the use of Gittins indices in the evaluation of strategies for families of alternative bandit processes has found many applications. Among these are procedures for sensitivity analysis in stochastic scheduling. This theoretical paper aims at developing results which will form the basis of an approach to strategy evaluation for a class of processes of greater complexity. These
... [Show full abstract] are Markov Decision Processes in parallel satisfying a condition first enunciated by Whittle. The theory of Gittins indices forms the basis of the analysis. Read more September 1998 · The Journal of Portfolio Management
The focus of this study is the effect that portfolio composition has on the tracking performance of indexed portfolios. Indexed portfolios from high-capitalization indices are shown to have a lower tracking error and standard deviation of tracking error than indexed portfolios from low-capitalization indices. Additionally, the results indicate that value-weighted indices can be indexed more
... [Show full abstract] accurately than equally-weighted indices. Moreover, stratification across industry groups does not appear to be as important for forming a low-capitalization indexed portfolio as it is for forming a high-capitalization indexed portfolio. Finally, the rebalancing frequency of solving for the optimal portfolio weights for the stocks in the indexed portfolios does not appear to be critical factor affecting tracking performance. Read more April 2012
The Network Voronoi diagram and its variants have been extensively used in the context of numerous applications in road networks, particularly to efficiently evaluate various spatial proximity queries such as k nearest neighbor (kNN), reverse kNN, and closest pair. Although the existing approaches successfully utilize the network Voronoi diagram as a way to partition the space for their specific
... [Show full abstract] problems, there is little emphasis on how to efficiently find and access the network Voronoi cell containing a particular point or edge of the network. In this paper, we study the index structures on network Voronoi diagrams that enable exact and fast response to contain query in road networks. We show that existing index structures, treating a network Voronoi cell as a simple polygon, may yield inaccurate results due to the network topology, and fail to scale to large networks with numerous Voronoi generators. With our method, termed Voronoi-Quad-tree (or VQ-tree for short), we use Quad-tree to index network Voronoi diagrams to address both of these shortcomings. We demonstrate the efficiency of VQ-tree via experimental evaluations with real-world datasets consisting of a variety of large road networks with numerous data objects. Read more June 1987
Sometimes it is required to output a block of word data from an 8-bit processor to a 16-bit system. Usually, data bytes are written alternately to two ports whose outputs are used together as one 16-bit port. This process may be performed faster using the 16-bit address bus as a data bus as indicated. An arrangement is described using the 6800 processor as an example, and a routine listing is
... [Show full abstract] tabulated. An indirect indexing operation is used to place each dataword first into the index register and then onto the address bus. Read more January 2009 · Fundamenta Informaticae
This paper describes the idea of a multi-dimensional bucket index designed for efficient indexing of telemetric readings. This data structure answers spatio-temporal range queries concerning utility usage within user selected region and time. In addition, it has a capability to adjust to incoming data and therefore is suitable to process data of highly dynamic nature. The paper also presents a
... [Show full abstract] stochastic prediction method to estimate utility usage in the near future. Read more Article
Full-text available
January 2014 · Advances in Intelligent Systems and Computing
High-dimensional indexes do not work because of the often-cited "curse of dimensionality." However, users are usually interested in querying data over a relatively small subset of the entire attribute set at a time. A potential solution is to use lower dimensional indexes that accurately represent the user access patterns. To address these issues, in this paper we propose a parameterizable
... [Show full abstract] technique to recommend indexes based on index types. View full-text January 2005
This paper addresses the problem of splitting trajectories optimally for the purpose of efficiently supporting spatio-temporal range queries using index structures (e.g., R-trees) that use minimum bounding hyper-rectangles as trajectory approximations. We derive a formal cost model for estimating the number of I/Os required to evaluate a spatio-temporal range query with respect to a given query
... [Show full abstract] size and an arbitrary split of a trajectory. Based on the proposed model, we introduce a dynamic programming algorithm for splitting a set of trajectories that minimizes the number of expected disk I/Os with respect to an average query size. In addition, we develop a linear time, near optimal solution for this problem to be used in a dynamic case where trajectory points are continuously updated. Our experimental evaluation confirms the effectiveness and efficiency of our proposed splitting policies when embedded into an R-tree structure. Read more December 1988 · Reference Services Review
When I began as column editor for “Reference Serials” I experienced an immediate need for a cumulative index to titles reviewed in this column over the years. An expeditious solution to the question of whether a particular title had been reviewed previously in RSR was a search of Book Review Index on DIALOG. In the process of this searching, I discovered that BRI's indexing of RSR was more
... [Show full abstract] complete than RSR's self indexing and, of course, BRI's indexing online is cumulative while RSR has no cumulative indexes. Both index options had an additional problem: each included references to all the titles mentioned in the various columns of RSR regardless of whether the reference was actually to an extensive review or simply to an unannotated listing of the title. Read more July 2019
We consider the Wiener integral with respect to a $d$-parameter Hermite process with Hurst multi-index ${\bf H}= (H_{1},\ldots, H_{d}) \in \left( \frac{1}{2}, 1\right) ^{d}$ and we analyze the limit behavior in distribution of this object when the components of ${\bf H}$ tend to $1$ and/or $\frac{1}{2}$. As examples, we focus on the solution to the stochastic heat equation with additive Hermite
... [Show full abstract] noise and to the Hermite Ornstein-Uhlenbeck process. Read more Conference Paper
Full-text available
January 2011
This paper presents an evaluation of the mqr-tree for indexing a database of line data. Many spatial access methods have been proposed for handling either point or region data, with the vast majority able to handle these data types efficiently. However, line segment data presents challenges for most spatial access methods. Recent work on the mqr-tree showed much potential for efficiently indexing
... [Show full abstract] line data. We identify limitations of the data sets in the initial evaluation. Then, we further evaluate the ability of the mqr-tree to efficiently index line sets with different organizations that address the limitations of the initial test. A comparison versus the R-tree shows that the mqr-tree achieves significantly lower overlap and overcoverage, which makes the mqr-tree a significant candidate for indexing line and line-segment data. View full-text January 2002 · The VLDB Journal
Real-world entities are inherently spatially and temporally referenced, and database applications increasingly exploit databases
that record the past, present, and anticipated future locations of entities, e.g., the residences of customers obtained by
the geo-coding of addresses. Indices that efficiently support queries on the spatio-temporal extents of such entities are
needed. However, past
... [Show full abstract] indexing research has progressed in largely separate spatial and temporal streams. Adding time dimensions
to spatial indices, as if time were a spatial dimension, neither supports nor exploits the special properties of time. On
the other hand, temporal indices are generally not amenable to extension with spatial dimensions. This paper proposes the
first efficient and versatile index for a general class of spatio-temporal data: the discretely changing spatial aspect of
an object may be a point or may have an extent; both transaction time and valid time are supported, and a generalized notion
of the current time, now, is accommodated for both temporal dimensions. The index is based on the R-tree and provides means of prioritizing space versus time, which enables it to adapt to spatially and temporally restrictive
queries. Performance experiments are reported that evaluate pertinent aspects of the index. Read more Last Updated: 05 Jul 2022
Looking for the full-text? You can request the full-text of this chapter directly from the authors on ResearchGate.
Request full-text
