Search Constraints
1  9 of 9
Number of results to display per page
Search Results

 Resource Type:
 Conference Proceeding
 Creator:
 Oommen, B. John and Polk, Spencer
 Abstract:
 The field of game playing is a particularly wellstudied area within the context of AI, leading to the development of powerful techniques, such as the alphabeta search, capable of achieving competitive game play against an intelligent opponent. It is well known that tree pruning strategies, such as alphabeta, benefit strongly from proper move ordering, that is, searching the best element first. Inspired by the formerly unrelated field of Adaptive Data Structures (ADSs), we have previously introduced the HistoryADS technique, which employs an adaptive list to achieve effective and dynamic move ordering, in a domain independent fashion, and found that it performs well in a wide range of cases. However, previous work did not compare the performance of the HistoryADS heuristic to any established move ordering strategy. In an attempt to address this problem, we present here a comparison to two wellknown, acclaimed strategies, which operate on a similar philosophy to the HistoryADS, the History Heuristic, and the Killer Moves technique. We find that, in a wide range of twoplayer and multiplayer games, at various points in the game’s progression, the HistoryADS performs at least as well as these strategies, and, in fact, outperforms them in the majority of cases.
 Date Created:
 20160101

 Resource Type:
 Conference Proceeding
 Creator:
 Oommen, B. John and Astudillo, César A.
 Abstract:
 We present a method that employs a treebased Neural Network (NN) for performing classification. The novel mechanism, apart from incorporating the information provided by unlabeled and labeled instances, rearranges the nodes of the tree as per the laws of Adaptive Data Structures (ADSs). Particularly, we investigate the Pattern Recognition (PR) capabilities of the TreeBased TopologyOriented SOM (TTOSOM) when Conditional Rotations (CONROT) [8] are incorporated into the learning scheme. The learning methodology inherits all the properties of the TTOSOMbased classifier designed in [4]. However, we now augment it with the property that frequently accessed nodes are moved closer to the root of the tree. Our experimental results show that on average, the classification capabilities of our proposed strategy are reasonably comparable to those obtained by some of the stateoftheart classification schemes that only use labeled instances during the training phase. The experiments also show that improved levels of accuracy can be obtained by imposing trees with a larger number of nodes.
 Date Created:
 20150101

 Resource Type:
 Conference Proceeding
 Creator:
 Tavasoli, Hanane, Oommen, B. John, and Yazidi, Anis
 Abstract:
 In this paper, we propose a novel online classifier for complex data streams which are generated from nonstationary stochastic properties. Instead of using a single training model and counters to keep important data statistics, the introduced online classifier scheme provides a realtime selfadjusting learning model. The learning model utilizes the multiplicationbased update algorithm of the Stochastic Learning Weak Estimator (SLWE) at each time instant as a new labeled instance arrives. In this way, the data statistics are updated every time a new element is inserted, without requiring that we have to rebuild its model when changes occur in the data distributions. Finally, and most importantly, the model operates with the understanding that the correct classes of previouslyclassified patterns become available at a later juncture subsequent to some time instances, thus requiring us to update the training set and the training model. The results obtained from rigorous empirical analysis on multinomial distributions, is remarkable. Indeed, it demonstrates the applicability of our method on synthetic datasets, and proves the advantages of the introduced scheme.
 Date Created:
 20160101

 Resource Type:
 Conference Proceeding
 Creator:
 Yazidi, Anis, Oommen, B. John, and Hammer, Hugo Lewi
 Abstract:
 The problem of clustering, or unsupervised classification, has been solved by a myriad of techniques, all of which depend, either directly or implicitly, on the Bayesian principle of optimal classification. To be more specific, within a Bayesian paradigm, if one is to compare the testing sample with only a single point in the feature space from each class, the optimal Bayesian strategy would be to achieve this based on the distance from the corresponding means or central points in the respective distributions. When this principle is applied in clustering, one would assign an unassigned sample into the cluster whose mean is the closest, and this can be done in either a bottomup or a topdown manner. This paper pioneers a clustering achieved in an “AntiBayesian” manner, and is based on the breakthrough classification paradigm pioneered by Oommen et al. The latter relies on a radically different approach for classifying data points based on the noncentral quantiles of the distributions. Surprisingly and counterintuitively, this turns out to work equally or closetoequally well to an optimal supervised Bayesian scheme, which thus begs the natural extension to the unexplored arena of clustering. Our algorithm can be seen as the AntiBayesian counterpart of the wellknown kmeans algorithm (The fundamental AntiBayesian paradigm need not just be used to the kmeans principle. Rather, we hypothesize that it can be adapted to any of the scores of techniques that is indirectly based on the Bayesian paradigm.), where we assign points to clusters using quantiles rather than the clusters’ centroids. Extensive experimentation (This paper contains the prima facie results of experiments done on one and twodimensional data. The extensions to multidimensional data are not included in the interest of space, and would use the corresponding multidimensional AntiNa¨ıveBayes classification rules given in [1].) demonstrates that our AntiBayesian clustering converges fast and with precision results competitive to a kmeans clustering.
 Date Created:
 20150101

 Resource Type:
 Conference Proceeding
 Creator:
 Polk, Spencer and Oommen, B. John
 Abstract:
 This paper pioneers the avenue of enhancing a wellknown paradigm in game playing, namely the use of Historybased heuristics, with a totallyunrelated area of computer science, the field of Adaptive Data Structures (ADSs). It is a wellknown fact that highlyregarded game playing strategies, such as alphabeta search, benefit strongly from proper move ordering, and from this perspective, the History heuristic is, probably, one of the most acclaimed techniques used to achieve AIbased game playing. Recently, the authors of this present paper have shown that techniques derived from the field of ADSs, which are concerned with query optimization in a data structure, can be applied to move ordering in multiplayer games. This was accomplished by ranking opponent threat levels. The work presented in this paper seeks to extend the utility of ADSbased techniques to twoplayer and multiplayer games, through the development of a new move ordering strategy that incorporates the historical advantages of the moves. The resultant technique, the HistoryADS heuristic, has been found to produce substantial (i.e, even up to 70%) savings in a variety of twoplayer and multiplayer games, at varying ply depths, and at both initial and midgame board states. As far as we know, results of this nature have not been reported in the literature before.
 Date Created:
 20150101

 Resource Type:
 Conference Proceeding
 Creator:
 Oommen, B. John and Kim, SangWoon
 Abstract:
 This paper deals with the relatively new field of sequencebased estimation which involves utilizing both the information in the observations and in their sequence of appearance. Our intention is to obtain Maximum Likelihood estimates by “extracting” the information contained in the observations when perceived as a sequence rather than as a set. The results of [15] introduced the concepts of Sequence Based Estimation (SBE) for the Binomial distribution. This current paper generalizes these results for the multinomial “twoatatime” scenario. We invoke a novel phenomenon called “Occlusion” that can be described as follows: By “concealing” certain observations, we map the estimation problem onto a lowerdimensional binomial space. Once these occluded SBEs have been computed, we demonstrate how the overall Multinomial SBE (MSBE) can be obtained by mapping several lowerdimensional estimates onto the original higherdimensional space. We formally prove and experimentally demonstrate the convergence of the corresponding estimates.
 Date Created:
 20160101

 Resource Type:
 Conference Proceeding
 Creator:
 Kim, SangWoon and Oommen, B. John
 Abstract:
 The Maximum Likelihood (ML) and Bayesian estimation paradigms work within the model that the data, from which the parameters are to be estimated, is treated as a set rather than as a sequence. The pioneering paper that dealt with the field of sequencebased estimation [2] involved utilizing both the information in the observations and in their sequence of appearance. The results of [2] introduced the concepts of Sequence Based Estimation (SBE) for the Binomial distribution, where the authors derived the corresponding MLE results when the samples are taken twoatatime, and then extended these for the cases when they are processed threeatatime, fouratatime etc. These results were generalized for the multinomial “twoatatime” scenario in [3]. This paper (This paper is dedicated to the memory of Dr. Mohamed Kamel, who was a close friend of the first author.) now further generalizes the results found in [3] for the multinomial case and for subsequences of length 3. The strategy used in [3] (and also here) involves a novel phenomenon called “Occlusion” that has not been reported in the field of estimation. The phenomenon can be described as follows: By occluding (hiding or concealing) certain observations, we map the estimation problem onto a lowerdimensional space, i.e., onto a binomial space. Once these occluded SBEs have been computed, the overall Multinomial SBE (MSBE) can be obtained by combining these lowerdimensional estimates. In each case, we formally prove and experimentally demonstrate the convergence of the corresponding estimates.
 Date Created:
 20160101

 Resource Type:
 Conference Proceeding
 Creator:
 Loke, R. K.S. and Oommen, B. John
 Abstract:
 We consider a problem which can greatly enhance the areas of cursive script recognition and the recognition of printed character sequences. This problem involves recognizing words/strings by processing their noisy subsequences. Let X* be any unknown word from a finite dictionary H. Let U be any arbitrary subsequence of X*. We study the problem of estimating X* by processing Y, a noisy version of U. Y contains substitution, insertion, deletion and generalized transposition errors — the latter occurring when transposed characters are themselves subsequently substituted. We solve the noisy subsequence recognition problem by defining and using the constrained edit distance between X ε H and Y subject to any arbitrary edit constraint involving the number and type of edit operations to be performed. An algorithm to compute this constrained edit distance has been presented. Using these algorithms we present a syntactic Pattern Recognition (PR) scheme which corrects noisy text containing all these types of errors. Experimental results which involve strings of lengths between 40 and 80 with an average of 30.24 deleted characters and an overall average noise of 68.69 % demonstrate the superiority of our system over existing methods.
 Date Created:
 19950101

 Resource Type:
 Conference Proceeding
 Creator:
 Oommen, B. John, Zhan, Justin, and Crisostomo, Johanna
 Abstract:
 Anomaly detection involves identifying observations that deviate from the normal behavior of a system. One of the ways to achieve this is by identifying the phenomena that characterize "normal" observations. Subsequently, based on the characteristics of data learned from the normal observations, new observations are classified as being either normal or not. Most stateoftheart approaches, especially those which belong to the family parameterized statistical schemes, work under the assumption that the underlying distributions of the observations are stationary. That is, they assume that the distributions that are learned during the training (or learning) phase, though unknown, are not timevarying. They further assume that the same distributions are relevant even as new observations are encountered. Although such a " stationarity" assumption is relevant for many applications, there are some anomaly detection problems where stationarity cannot be assumed. For example, in network monitoring, the patterns which are learned to represent normal behavior may change over time due to several factors such as network infrastructure expansion, new services, growth of user population, etc. Similarly, in meteorology, identifying anomalous temperature patterns involves taking into account seasonal changes of normal observations. Detecting anomalies or outliers under these circumstances introduces several challenges. Indeed, the ability to adapt to changes in nonstationary environments is necessary so that anomalous observations can be identified even with changes in what would otherwise be classified as normal behavior. In this paper, we proposed to apply weak estimation theory for anomaly detection in dynamic environments. In particular, we apply this theory to detect anomaly activities in system calls. Our experimental results demonstrate that our proposal is both feasible and effective for the detection of such anomalous activities.
 Date Created:
 20120922