Dhananjay Raghavan Thiruvady

Functional connectivity of the prefrontal cortex in Huntington’s disease

Background: Huntington’s disease is a progressive neurodegenerative disorder that results in deterioration and atrophy of various brain regions. Aim: To assess the functional connectivity between prefrontal brain regions in patients with Huntington’s disease, compared with normal controls, using functional magnetic resonance imaging. Patients and methods: 20 patients with Huntington’s disease and 17 matched controls performed a Simon task that is known to activate lateral prefrontal and anterior cingulate cortical regions. The functional connectivity was hypothesised to be impaired in patients with Huntington’s disease between prefrontal regions of interest, selected from both hemispheres, in the anterior cingulate and dorsal lateral prefrontal cortex. Results: Controls showed a dynamic increase in interhemispheric functional connectivity during task performance, compared with the baseline state; patients with Huntington’s disease, however, showed no such increase in prefrontal connectivity. Overall, patients with Huntington’s disease showed significantly impaired functional connectivity between anterior cingulate and lateral prefrontal regions in both hemispheres compared with controls. Furthermore, poor task performance was predicted by reduced connectivity in patients with Huntington’s disease between the left anterior cingulate and prefrontal regions. Conclusions: This finding represents a loss of synchrony in activity between prefrontal regions in patients with Huntington’s disease when engaged in the task, which predicted poor task performance. Results show that functional interactions between critical prefrontal regions, necessary for cognitive performance, are compromised in Huntington’s disease. It is speculated whether significantly greater levels of activation in patients with Huntington’s disease (compared with controls) observed in several brain regions partially compensate for the otherwise compromised interactions between cortical regions.

Increased cortical recruitment in Huntington's disease using a Simon task.

Cognitive deficits in Huntingtons disease (HD) have been attributed to neuronal degeneration within the striatum; however, postmortem and structural imaging studies have revealed more widespread morphological changes. To examine the impact of HD-related changes in regions outside the striatum, we used functional magnetic resonance imaging (fMRI) in HD to examine brain activation patterns using a Simon task that required a button press response to either congruent or incongruent arrow stimuli. Twenty mild to moderate stage HD patients and 17 healthy controls were scanned using a 3T GE scanner. Data analysis involved the use of statistical parametric mapping software with a random effects analysis model to investigate group differences brain activation patterns compared to baseline. HD patients recruited frontal and parietal cortical regions to perform the task, and also showed significantly greater activation, compared to controls, in the caudal anterior cingulate, insula, inferior parietal lobules, superior temporal gyrus bilaterally, right inferior frontal gyrus, right precuneus/superior parietal lobule, left precentral gyrus, and left dorsal premotor cortex. The significantly increased activation in anterior cingulate-frontal-motor-parietal cortex in HD may represent a primary dysfunction due to direct cell loss or damage in cortical regions, and/or a secondary compensatory mechanism of increased cortical recruitment due to primary striatal deficits.

Mining Negative Rules Using GRD

GRD is an algorithm for k-most interesting rule discovery. In contrast to association rule discovery, GRD does not require the use of a minimum support constraint. Rather, the user must specify a measure of interestingness and the number of rules sought (k). This paper reports efficient techniques to extend GRD to support mining of negative rules. We demonstrate that the new approach provides tractable discovery of both negative and positive rules.

Parallel ant colony optimization for resource constrained job scheduling

In mining supply chains, large combinatorial optimization problems arise. These are NP-hard and typically require a large number of computing resources to solve them. In particular, the run-time overheads can become increasingly prohibitive with increasing problem sizes. Parallel methods provide a way to manage such run-time issues by utilising several processors in independent or shared memory architectures. However it is not obvious how to adapt serial optimisation algorithms to perform best in a parallel environment. Here, we consider a resource constrained scheduling problem which is motivated in mining supply chains and present two popular meta-heuristics, ant colony optimization (ACO) and simulated annealing and investigate how best to parallelize these methods on a shared memory architecture consisting of several cores. ACO’s solution construction framework is inherently parallel allowing a relatively straightforward parallel implementation. However, for best performance, ACO needs an element of local search. This significantly complicates the paralellization. Several alternative schemes for parallel ACO with elements of local search are considered and evaluated empirically. We find that ACO with local search is the most effective single-threaded algorithm. The best parallel implementation can obtain similar quality results to the serial method in significantly less elapsed time.

Hybridizing Beam-ACO with Constraint Programming for Single Machine Job Scheduling

A recent line of research concerns the integration of ant colony optimization and constraint programming. Hereby, constraint programming is used for eliminating parts of the search tree during the solution construction of ant colony optimization. In the context of a single machine scheduling problem, for example, it has been shown that the integration of constraint programming can significantly improve the ability of ant colony optimization to find feasible solutions. One of the remaining problems, however, concerns the elevated computation time requirements of the hybrid algorithm, which are due to constraint propagation. In this work we propose a possible solution to this problem by integrating constraint programming with a specific version of ant colony optimization known as Beam-ACO. The idea is to reduce the time spent for constraint propagation by parallelizing the solution construction process as done in Beam-ACO. The results of the proposed algorithm show indeed that it is currently the best performing algorithm for the above mentioned single machine job scheduling problem.

Car sequencing with constraint-based ACO

Hybrid methods for solving combinatorial optimization problems have become increasingly popular recently. The present paper is concerned with hybrids of ant colony optimization and constraint programming which are typically useful for problems with hard constraints. However, the original algorithm suffered from large CPU time requirements. It was shown that such an integration can be made efficient via a further hybridization with beam search resulting in CP-Beam-ACO. The original work suggested this in the context of job scheduling. We show here that this algorithm type is also effective on another problem class, namely the car sequencing. We consider an optimization version, where we aim to optimize the utilization rates across the sequence. Car sequencing is a notoriously difficult problem, because it is difficult to obtain good bounds via relaxations. We show that stochastic sampling provides superior results to well known lower bounds for this problem when combined with CP-Beam-ACO.

Constraint Programming and Ant Colony System for the Component Deployment Problem

Abstract Contemporary motor vehicles have increasing numbers of automated functions to augment the safety and comfort of a car. The automotive industry has to incorporate increasing numbers of processing units in the structure of cars to run the software that provides these functionalities. The software components often need access to sensors or mechanical devices which they are designed to operate. The result is a network of hardware units which can accommodate a limited number of software programs, each of which has to be assigned to a hardware unit. A prime goal of this deployment problem is to find software-to-hardware assignments that maximise the reliability of the system. In doing so, the assignments have to observe a number of constraints to be viable. This includes limited memory of a hardware unit, collocation of software components on the same hardware units, and communication between software components. Since the problem consists of many constraints with a significantly large search space, we investigate an ACO and constraint programming (CP) hybrid for this problem. We find that despite the large number of constraints, ACO on its own is the most effective method providing good solutions by also exploring infeasible regions.

Hybrids of Integer Programming and ACO for Resource Constrained Job Scheduling

A recent line of research considers hybrids of Lagrangian relaxation and Ant Colony Optimisation (ACO). Studies have shown that for hard constrained optimisation problems Lagrangian relaxation can effectively guide ACO to provide good feasible solutions. We consider applying these ideas to create a matheuristic combining ACO with decomposition approaches from mathematical programming for a resource constrained job scheduling problem. We are given a number of jobs which have to be executed on a number of machines satisfying several constraints. These include precedences and release times within machines and the machines are linked via a central resource constraint. By removing the linking constraint, the each machine’s scheduling problem can be solved independently as a relatively simple subproblem. Both Danzig-Wolfe decomposition with column generation and Lagrangian relaxation are tried to carry out this decomposition. The relaxed solutions can provide useful guidance to determine solutions either via problem specific heuristics and ACO. Empirical results show that the Lagrangian relaxation matheuristic performs well in limited time-frames whereas the column generation based heuristic provides improved lower and upper bounds when run to convergence.

Strip packing with hybrid ACO: Placement order is learnable

This paper investigates the use of hybrid meta-heuristics based on ant colony optimization (ACO) for the strip packing problem. Here, a fixed set of rectangular items of fixed sizes have to be placed on a strip of fixed width and infinite height without overlaps and with the objective to minimize the height used. We analyze a commonly used basic placement heuristic (BLF) by itself and in a number of hybrid combinations with ACO. We compare versions that learn item order only, item rotation only, both independently, and rotations conditionally upon placement order. Our analysis shows that integrating a learning meta-heuristic provides a significant performance advantage over using the basic placement heuristic by itself. The experiments confirm that even just learning a placement order alone can provide significant performance improvements. Interestingly, learning item rotations provides at best a marginal advantage. The best hybrid algorithm presented in this paper significantly outperforms previously reported strip packing meta-heuristics.

A mixed integer linear programming model for reliability optimisation in the component deployment problem

Component deployment is a combinatorial optimisation problem in software engineering that aims at finding the best allocation of software components to hardware resources in order to optimise quality attributes, such as reliability. The problem is often constrained because of the limited hardware resources, and the communication network, which may connect only certain resources. Owing to the non-linear nature of the reliability function, current optimisation methods have focused mainly on heuristic or metaheuristic algorithms. These are approximate methods, which find near-optimal solutions in a reasonable amount of time. In this paper, we present a mixed integer linear programming (MILP) formulation of the component deployment problem. We design a set of experiments where we compare the MILP solver to methods previously used to solve this problem. Results show that the MILP solver is efficient in finding feasible solutions even where other methods fail, or prove infeasibility where feasible solutions do not exist.