On the Erdos-Szekeres n-interior point problem

The n-interior point variant of the Erdos-Szekeres problem is to show the following: For any n, n-1, every point set in the plane with sufficient number of interior points contains a convex polygon containing exactly n-interior points. This has been proved only for n-3. In this paper, we prove it for pointsets having atmost logarithmic number of convex layers. We also show that any pointset containing atleast n interior points, there exists a 2-convex polygon that contains exactly n-interior points.

A sequential dual method for structural SVMs

In many real world prediction problems the output is a structured object like a sequence or a tree or a graph. Such problems range from natural language processing to compu- tational biology or computer vision and have been tackled using algorithms, referred to as structured output learning algorithms. We consider the problem of structured classifi- cation. In the last few years, large margin classifiers like sup-port vector machines (SVMs) have shown much promise for structured output learning. The related optimization prob -lem is a convex quadratic program (QP) with a large num-ber of constraints, which makes the problem intractable for large data sets. This paper proposes a fast sequential dual method (SDM) for structural SVMs. The method makes re-peated passes over the training set and optimizes the dual variables associated with one example at a time. The use of additional heuristics makes the proposed method more efficient. We present an extensive empirical evaluation of the proposed method on several sequence learning problems.Our experiments on large data sets demonstrate that the proposed method is an order of magnitude faster than state of the art methods like cutting-plane method and stochastic gradient descent method (SGD). Further, SDM reaches steady state generalization performance faster than the SGD method. The proposed SDM is thus a useful alternative for large scale structured output learning.

A game theoretic approach for feature clustering and its application to feature selection

In this paper, we develop a game theoretic approach for clustering features in a learning problem. Feature clustering can serve as an important preprocessing step in many problems such as feature selection, dimensionality reduction, etc. In this approach, we view features as rational players of a coalitional game where they form coalitions (or clusters) among themselves in order to maximize their individual payoffs. We show how Nash Stable Partition (NSP), a well known concept in the coalitional game theory, provides a natural way of clustering features. Through this approach, one can obtain some desirable properties of the clusters by choosing appropriate payoff functions. For a small number of features, the NSP based clustering can be found by solving an integer linear program (ILP). However, for large number of features, the ILP based approach does not scale well and hence we propose a hierarchical approach. Interestingly, a key result that we prove on the equivalence between a k-size NSP of a coalitional game and minimum k-cut of an appropriately constructed graph comes in handy for large scale problems. In this paper, we use feature selection problem (in a classification setting) as a running example to illustrate our approach. We conduct experiments to illustrate the efficacy of our approach.

A constant factor approximation algorithm for boxicity of circular arc graphs

Boxicity of a graph G(V, E) is the minimum integer k such that G can be represented as the intersection graph of k-dimensional axis parallel boxes in Rk. Equivalently, it is the minimum number of interval graphs on the vertex set V such that the intersection of their edge sets is E. It is known that boxicity cannot be approximated even for graph classes like bipartite, co-bipartite and split graphs below O(n0.5-ε)-factor, for any ε > 0 in polynomial time unless NP = ZPP. Till date, there is no well known graph class of unbounded boxicity for which even an nε-factor approximation algorithm for computing boxicity is known, for any ε < 1. In this paper, we study the boxicity problem on Circular Arc graphs - intersection graphs of arcs of a circle. We give a (2+ 1/k)-factor polynomial time approximation algorithm for computing the boxicity of any circular arc graph along with a corresponding box representation, where k ≥ 1 is its boxicity. For Normal Circular Arc(NCA) graphs, with an NCA model given, this can be improved to an additive 2-factor approximation algorithm. The time complexity of the algorithms to approximately compute the boxicity is O(mn+n2) in both these cases and in O(mn+kn2) which is at most O(n3) time we also get their corresponding box representations, where n is the number of vertices of the graph and m is its number of edges. The additive 2-factor algorithm directly works for any Proper Circular Arc graph, since computing an NCA model for it can be done in polynomial time.

Cubicity, degeneracy, and crossing number

A $k$-box $B=(R_1,...,R_k)$, where each $R_i$ is a closed interval on the real line, is defined to be the Cartesian product $R_1\times R_2\times ...\times R_k$. If each $R_i$ is a unit length interval, we call $B$ a $k$-cube. Boxicity of a graph $G$, denoted as $\boxi(G)$, is the minimum integer $k$ such that $G$ is an intersection graph of $k$-boxes. Similarly, the cubicity of $G$, denoted as $\cubi(G)$, is the minimum integer $k$ such that $G$ is an intersection graph of $k$-cubes. It was shown in [L. Sunil Chandran, Mathew C. Francis, and Naveen Sivadasan: Representing graphs as the intersection of axis-parallel cubes. MCDES-2008, IISc Centenary Conference, available at CoRR, abs/cs/ 0607092, 2006.] that, for a graph $G$ with maximum degree $\Delta$, $\cubi(G)\leq \lceil 4(\Delta +1)\log n\rceil$. In this paper, we show that, for a $k$-degenerate graph $G$, $\cubi(G) \leq (k+2) \lceil 2e \log n \rceil$. Since $k$ is at most $\Delta$ and can be much lower, this clearly is a stronger result. This bound is tight. We also give an efficient deterministic algorithm that runs in $O(n^2k)$ time to output a $8k(\lceil 2.42 \log n\rceil + 1)$ dimensional cube representation for $G$. An important consequence of the above result is that if the crossing number of a graph $G$ is $t$, then $\boxi(G)$ is $O(t^{1/4}{\lceil\log t\rceil}^{3/4})$ . This bound is tight up to a factor of $O((\log t)^{1/4})$. We also show that, if $G$ has $n$ vertices, then $\cubi(G)$ is $O(\log n + t^{1/4}\log t)$. Using our bound for the cubicity of $k$-degenerate graphs we show that cubicity of almost all graphs in $\mathcal{G}(n,m)$ model is $O(d_{av}\log n)$, where $d_{av}$ denotes the average degree of the graph under consideration. model is O(davlogn).

Improving query focused summarization using look-ahead strategy

Query focused summarization is the task of producing a compressed text of original set of documents based on a query. Documents can be viewed as graph with sentences as nodes and edges can be added based on sentence similarity. Graph based ranking algorithms which use 'Biased random surfer model' like topic-sensitive LexRank have been successfully applied to query focused summarization. In these algorithms, random walk will be biased towards the sentences which contain query relevant words. Specifically, it is assumed that random surfer knows the query relevance score of the sentence to where he jumps. However, neighbourhood information of the sentence to where he jumps is completely ignored. In this paper, we propose look-ahead version of topic-sensitive LexRank. We assume that random surfer not only knows the query relevance of the sentence to where he jumps but he can also look N-step ahead from that sentence to find query relevance scores of future set of sentences. Using this look ahead information, we figure out the sentences which are indirectly related to the query by looking at number of hops to reach a sentence which has query relevant words. Then we make the random walk biased towards even to the indirect query relevant sentences along with the sentences which have query relevant words. Experimental results show 20.2% increase in ROUGE-2 score compared to topic-sensitive LexRank on DUC 2007 data set. Further, our system outperforms best systems in DUC 2006 and results are comparable to state of the art systems.

Rainbow connectivity: hardness and tractability

A path in an edge colored graph is said to be a rainbow path if no two edges on the path have the same color. An edge colored graph is (strongly) rainbow connected if there exists a (geodesic) rainbow path between every pair of vertices. The (strong) rainbow connectivity of a graph G, denoted by (src(G), respectively) rc(G) is the smallest number of colors required to edge color the graph such that G is (strongly) rainbow connected. In this paper we study the rainbow connectivity problem and the strong rainbow connectivity problem from a computational point of view. Our main results can be summarised as below: 1) For every fixed k >= 3, it is NP-Complete to decide whether src(G) <= k even when the graph G is bipartite. 2) For every fixed odd k >= 3, it is NP-Complete to decide whether rc(G) <= k. This resolves one of the open problems posed by Chakraborty et al. (J. Comb. Opt., 2011) where they prove the hardness for the even case. 3) The following problem is fixed parameter tractable: Given a graph G, determine the maximum number of pairs of vertices that can be rainbow connected using two colors. 4) For a directed graph G, it is NP-Complete to decide whether rc(G) <= 2.

A hybrid RTS-BP algorithm for improved detection of large-MIMO M-QAM signals

Low-complexity near-optimal detection of large-MIMO signals has attracted recent research. Recently, we proposed a local neighborhood search algorithm, namely reactive tabu search (RTS) algorithm, as well as a factor-graph based belief propagation (BP) algorithm for low-complexity large-MIMO detection. The motivation for the present work arises from the following two observations on the above two algorithms: i) Although RTS achieved close to optimal performance for 4-QAM in large dimensions, significant performance improvement was still possible for higher-order QAM (e.g., 16-, 64-QAM). ii) BP also achieved near-optimal performance for large dimensions, but only for {±1} alphabet. In this paper, we improve the large-MIMO detection performance of higher-order QAM signals by using a hybrid algorithm that employs RTS and BP. In particular, motivated by the observation that when a detection error occurs at the RTS output, the least significant bits (LSB) of the symbols are mostly in error, we propose to first reconstruct and cancel the interference due to bits other than LSBs at the RTS output and feed the interference cancelled received signal to the BP algorithm to improve the reliability of the LSBs. The output of the BP is then fed back to RTS for the next iteration. Simulation results show that the proposed algorithm performs better than the RTS algorithm, and semi-definite relaxation (SDR) and Gaussian tree approximation (GTA) algorithms.

Automatic compilation of MATLAB programs for synergistic execution on heterogeneous processors

MATLAB is an array language, initially popular for rapid prototyping, but is now being increasingly used to develop production code for numerical and scientific applications. Typical MATLAB programs have abundant data parallelism. These programs also have control flow dominated scalar regions that have an impact on the program's execution time. Today's computer systems have tremendous computing power in the form of traditional CPU cores and throughput oriented accelerators such as graphics processing units(GPUs). Thus, an approach that maps the control flow dominated regions to the CPU and the data parallel regions to the GPU can significantly improve program performance. In this paper, we present the design and implementation of MEGHA, a compiler that automatically compiles MATLAB programs to enable synergistic execution on heterogeneous processors. Our solution is fully automated and does not require programmer input for identifying data parallel regions. We propose a set of compiler optimizations tailored for MATLAB. Our compiler identifies data parallel regions of the program and composes them into kernels. The problem of combining statements into kernels is formulated as a constrained graph clustering problem. Heuristics are presented to map identified kernels to either the CPU or GPU so that kernel execution on the CPU and the GPU happens synergistically and the amount of data transfer needed is minimized. In order to ensure required data movement for dependencies across basic blocks, we propose a data flow analysis and edge splitting strategy. Thus our compiler automatically handles composition of kernels, mapping of kernels to CPU and GPU, scheduling and insertion of required data transfer. The proposed compiler was implemented and experimental evaluation using a set of MATLAB benchmarks shows that our approach achieves a geometric mean speedup of 19.8X for data parallel benchmarks over native execution of MATLAB.

Analysing modifica-tions in the synthesis of multiple state mechanical devices using configuration space and topology graphs

Automated synthesis of mechanical designs is an important step towards the development of an intelligent CAD system. Research into methods for supporting conceptual design using automated synthesis has attracted much attention in the past decades. In our research, ten experimental studies are conducted to find out how designers synthesize solution concepts for multi-state mechanical devices. The designers are asked to think aloud, while carrying out the synthesis. These design synthesis processes are video recorded. It has been found that modification of kinematic pairs and mechanisms is the major activity carried out by all the designers. This paper presents an analysis of these synthesis processes using configuration space and topology graph to identify and classify the types of modifications that take place. Understanding of these modification processes and the context in which they happened is crucial to develop a system for supporting design synthesis of multiple state mechanical devices that is capable of creating a comprehensive variety of solution alternatives.

OTCYMIST: Otsu-Canny minimal spanning tree for born-digital images

Text segmentation and localization algorithms are proposed for the born-digital image dataset. Binarization and edge detection are separately carried out on the three colour planes of the image. Connected components (CC's) obtained from the binarized image are thresholded based on their area and aspect ratio. CC's which contain sufficient edge pixels are retained. A novel approach is presented, where the text components are represented as nodes of a graph. Nodes correspond to the centroids of the individual CC's. Long edges are broken from the minimum spanning tree of the graph. Pair wise height ratio is also used to remove likely non-text components. A new minimum spanning tree is created from the remaining nodes. Horizontal grouping is performed on the CC's to generate bounding boxes of text strings. Overlapping bounding boxes are removed using an overlap area threshold. Non-overlapping and minimally overlapping bounding boxes are used for text segmentation. Vertical splitting is applied to generate bounding boxes at the word level. The proposed method is applied on all the images of the test dataset and values of precision, recall and H-mean are obtained using different approaches.

Phase synchronization in brain networks derived from correlation between probabilities of recurrences in functional MRI data

It is increasingly being recognized that resting state brain connectivity derived from functional magnetic resonance imaging (fMRI) data is an important marker of brain function both in healthy and clinical populations. Though linear correlation has been extensively used to characterize brain connectivity, it is limited to detecting first order dependencies. In this study, we propose a framework where in phase synchronization (PS) between brain regions is characterized using a new metric ``correlation between probabilities of recurrence'' (CPR) and subsequent graph-theoretic analysis of the ensuing networks. We applied this method to resting state fMRI data obtained from human subjects with and without administration of propofol anesthetic. Our results showed decreased PS during anesthesia and a biologically more plausible community structure using CPR rather than linear correlation. We conclude that CPR provides an attractive nonparametric method for modeling interactions in brain networks as compared to standard correlation for obtaining physiologically meaningful insights about brain function.

Generalized distributive law for ML decoding of space-time block codes

The problem of designing good space-time block codes (STBCs) with low maximum-likelihood (ML) decoding complexity has gathered much attention in the literature. All the known low ML decoding complexity techniques utilize the same approach of exploiting either the multigroup decodable or the fast-decodable (conditionally multigroup decodable) structure of a code. We refer to this well-known technique of decoding STBCs as conditional ML (CML) decoding. In this paper, we introduce a new framework to construct ML decoders for STBCs based on the generalized distributive law (GDL) and the factor-graph-based sum-product algorithm. We say that an STBC is fast GDL decodable if the order of GDL decoding complexity of the code, with respect to the constellation size, is strictly less than M-lambda, where lambda is the number of independent symbols in the STBC. We give sufficient conditions for an STBC to admit fast GDL decoding, and show that both multigroup and conditionally multigroup decodable codes are fast GDL decodable. For any STBC, whether fast GDL decodable or not, we show that the GDL decoding complexity is strictly less than the CML decoding complexity. For instance, for any STBC obtained from cyclic division algebras which is not multigroup or conditionally multigroup decodable, the GDL decoder provides about 12 times reduction in complexity compared to the CML decoder. Similarly, for the Golden code, which is conditionally multigroup decodable, the GDL decoder is only half as complex as the CML decoder.

Performance improvement of short-length regular low-density parity-check codes with low-complexity post-processing

It is well known that extremely long low-density parity-check (LDPC) codes perform exceptionally well for error correction applications, short-length codes are preferable in practical applications. However, short-length LDPC codes suffer from performance degradation owing to graph-based impairments such as short cycles, trapping sets and stopping sets and so on in the bipartite graph of the LDPC matrix. In particular, performance degradation at moderate to high E-b/N-0 is caused by the oscillations in bit node a posteriori probabilities induced by short cycles and trapping sets in bipartite graphs. In this study, a computationally efficient algorithm is proposed to improve the performance of short-length LDPC codes at moderate to high E-b/N-0. This algorithm makes use of the information generated by the belief propagation (BP) algorithm in previous iterations before a decoding failure occurs. Using this information, a reliability-based estimation is performed on each bit node to supplement the BP algorithm. The proposed algorithm gives an appreciable coding gain as compared with BP decoding for LDPC codes of a code rate equal to or less than 1/2 rate coding. The coding gains are modest to significant in the case of optimised (for bipartite graph conditioning) regular LDPC codes, whereas the coding gains are huge in the case of unoptimised codes. Hence, this algorithm is useful for relaxing some stringent constraints on the graphical structure of the LDPC code and for developing hardware-friendly designs.

On the SIG-dimension of trees under the L (a)-metric

Let where be a set of points in d-dimensional space with a given metric rho. For a point let r (p) be the distance of p with respect to rho from its nearest neighbor in Let B(p,r (p) ) be the open ball with respect to rho centered at p and having the radius r (p) . We define the sphere-of-influence graph (SIG) of as the intersection graph of the family of sets Given a graph G, a set of points in d-dimensional space with the metric rho is called a d-dimensional SIG-representation of G, if G is isomorphic to the SIG of It is known that the absence of isolated vertices is a necessary and sufficient condition for a graph to have a SIG-representation under the L (a)-metric in some space of finite dimension. The SIG-dimension under the L (a)-metric of a graph G without isolated vertices is defined to be the minimum positive integer d such that G has a d-dimensional SIG-representation under the L (a)-metric. It is denoted by SIG (a)(G). We study the SIG-dimension of trees under the L (a)-metric and almost completely answer an open problem posed by Michael and Quint (Discrete Appl Math 127:447-460, 2003). Let T be a tree with at least two vertices. For each let leaf-degree(v) denote the number of neighbors of v that are leaves. We define the maximum leaf-degree as leaf-degree(x). Let leaf-degree{(v) = alpha}. If |S| = 1, we define beta(T) = alpha(T) - 1. Otherwise define beta(T) = alpha(T). We show that for a tree where beta = beta (T), provided beta is not of the form 2 (k) - 1, for some positive integer k a parts per thousand yen 1. If beta = 2 (k) - 1, then We show that both values are possible.