Focussed Crawling with large scale Ordinal Regression Solvers

In this paper we propose a novel, scalable, clustering based Ordinal Regression formulation, which is an instance of a Second Order Cone Program (SOCP) with one Second Order Cone (SOC) constraint. The main contribution of the paper is a fast algorithm, CB-OR, which solves the proposed formulation more eficiently than general purpose solvers. Another main contribution of the paper is to pose the problem of focused crawling as a large scale Ordinal Regression problem and solve using the proposed CB-OR. Focused crawling is an efficient mechanism for discovering resources of interest on the web. Posing the problem of focused crawling as an Ordinal Regression problem avoids the need for a negative class and topic hierarchy, which are the main drawbacks of the existing focused crawling methods. Experiments on large synthetic and benchmark datasets show the scalability of CB-OR. Experiments also show that the proposed focused crawler outperforms the state-of-the-art.

Clustering Based Large Margin Classification: A Scalable Approach using SOCP Formulation

This paper presents a novel Second Order Cone Programming (SOCP) formulation for large scale binary classification tasks. Assuming that the class conditional densities are mixture distributions, where each component of the mixture has a spherical covariance, the second order statistics of the components can be estimated efficiently using clustering algorithms like BIRCH. For each cluster, the second order moments are used to derive a second order cone constraint via a Chebyshev-Cantelli inequality. This constraint ensures that any data point in the cluster is classified correctly with a high probability. This leads to a large margin SOCP formulation whose size depends on the number of clusters rather than the number of training data points. Hence, the proposed formulation scales well for large datasets when compared to the state-of-the-art classifiers, Support Vector Machines (SVMs). Experiments on real world and synthetic datasets show that the proposed algorithm outperforms SVM solvers in terms of training time and achieves similar accuracies.

Large-Scale Elastic Net Regularized Linear Classification SVMs and Logistic Regression

Elastic Net Regularizers have shown much promise in designing sparse classifiers for linear classification. In this work, we propose an alternating optimization approach to solve the dual problems of elastic net regularized linear classification Support Vector Machines (SVMs) and logistic regression (LR). One of the sub-problems turns out to be a simple projection. The other sub-problem can be solved using dual coordinate descent methods developed for non-sparse L2-regularized linear SVMs and LR, without altering their iteration complexity and convergence properties. Experiments on very large datasets indicate that the proposed dual coordinate descent - projection (DCD-P) methods are fast and achieve comparable generalization performance after the first pass through the data, with extremely sparse models.

Prototype Selection on Large and Streaming Data

Since streaming data keeps coming continuously as an ordered sequence, massive amounts of data is created. A big challenge in handling data streams is the limitation of time and space. Prototype selection on streaming data requires the prototypes to be updated in an incremental manner as new data comes in. We propose an incremental algorithm for prototype selection. This algorithm can also be used to handle very large datasets. Results have been presented on a number of large datasets and our method is compared to an existing algorithm for streaming data. Our algorithm saves time and the prototypes selected gives good classification accuracy.

Adaptive Scalable SVD Unit for Fast Processing of Large LSE Problems

Singular Value Decomposition (SVD) is a key linear algebraic operation in many scientific and engineering applications. In particular, many computational intelligence systems rely on machine learning methods involving high dimensionality datasets that have to be fast processed for real-time adaptability. In this paper we describe a practical FPGA (Field Programmable Gate Array) implementation of a SVD processor for accelerating the solution of large LSE problems. The design approach has been comprehensive, from the algorithmic refinement to the numerical analysis to the customization for an efficient hardware realization. The processing scheme rests on an adaptive vector rotation evaluator for error regularization that enhances convergence speed with no penalty on the solution accuracy. The proposed architecture, which follows a data transfer scheme, is scalable and based on the interconnection of simple rotations units, which allows for a trade-off between occupied area and processing acceleration in the final implementation. This permits the SVD processor to be implemented both on low-cost and highend FPGAs, according to the final application requirements.

SiGMa: Simple Greedy Matching for Aligning Large Knowledge Bases

The Internet has enabled the creation of a growing number of large-scale knowledge bases in a variety of domains containing complementary information. Tools for automatically aligning these knowledge bases would make it possible to unify many sources of structured knowledge and answer complex queries. However, the efficient alignment of large-scale knowledge bases still poses a considerable challenge. Here, we present Simple Greedy Matching (SiGMa), a simple algorithm for aligning knowledge bases with millions of entities and facts. SiGMa is an iterative propagation algorithm which leverages both the structural information from the relationship graph as well as flexible similarity measures between entity properties in a greedy local search, thus making it scalable. Despite its greedy nature, our experiments indicate that SiGMa can efficiently match some of the world's largest knowledge bases with high precision. We provide additional experiments on benchmark datasets which demonstrate that SiGMa can outperform state-of-the-art approaches both in accuracy and efficiency.

Large-Scale Neural Systems for Vision and Cognition

— Consideration of how people respond to the question What is this? has suggested new problem frontiers for pattern recognition and information fusion, as well as neural systems that embody the cognitive transformation of declarative information into relational knowledge. In contrast to traditional classification methods, which aim to find the single correct label for each exemplar (This is a car), the new approach discovers rules that embody coherent relationships among labels which would otherwise appear contradictory to a learning system (This is a car, that is a vehicle, over there is a sedan). This talk will describe how an individual who experiences exemplars in real time, with each exemplar trained on at most one category label, can autonomously discover a hierarchy of cognitive rules, thereby converting local information into global knowledge. Computational examples are based on the observation that sensors working at different times, locations, and spatial scales, and experts with different goals, languages, and situations, may produce apparently inconsistent image labels, which are reconciled by implicit underlying relationships that the network’s learning process discovers. The ARTMAP information fusion system can, moreover, integrate multiple separate knowledge hierarchies, by fusing independent domains into a unified structure. In the process, the system discovers cross-domain rules, inferring multilevel relationships among groups of output classes, without any supervised labeling of these relationships. In order to self-organize its expert system, the ARTMAP information fusion network features distributed code representations which exploit the model’s intrinsic capacity for one-to-many learning (This is a car and a vehicle and a sedan) as well as many-to-one learning (Each of those vehicles is a car). Fusion system software, testbed datasets, and articles are available from http://cns.bu.edu/techlab.

Anatomical identification of extracellularly recorded cells in large-scale multielectrode recordings.

This study combines for the first time two major approaches to understanding the function and structure of neural circuits: large-scale multielectrode recordings, and confocal imaging of labeled neurons. To achieve this end, we develop a novel approach to the central problem of anatomically identifying recorded cells, based on the electrical image: the spatiotemporal pattern of voltage deflections induced by spikes on a large-scale, high-density multielectrode array. Recordings were performed from identified ganglion cell types in the macaque retina. Anatomical images of cells in the same preparation were obtained using virally transfected fluorescent labeling or by immunolabeling after fixation. The electrical image was then used to locate recorded cell somas, axon initial segments, and axon trajectories, and these signatures were used to identify recorded cells. Comparison of anatomical and physiological measurements permitted visualization and physiological characterization of numerically dominant ganglion cell types with high efficiency in a single preparation.

Isolation of retinal progenitor and stem cells from the porcine eye

Purpose: Retinal progenitor cells (RPCs) and retinal stem cells (RSCs) from rodents and humans have been isolated and characterized in vitro. Transplantation experiments have confirmed their potential as tools for cell replacement in retinal degenerative diseases. The pig represents an ideal pre-clinical animal model to study the impact of transplantation because of the similarity of its eye to the human eye. However, little is known about porcine RPCs and RSCs. We aimed to identify and characterize in vitro RPCs and RSCs from porcine ocular tissues. Methods: Cells from different subregions of embryonic, postnatal and adult porcine eyes were grown in suspension sphere culture in serum-free medium containing basic fibroblast growth factor (bFGF) and epidermal growth factor (EGF). Growth curves and BrdU incorporation assays were performed to establish the proliferative capacity of isolated porcine retina-derived RPCs and ciliary epithelium (CE)-derived RSCs. Self-renewal potential was investigated by subsphere formation assays. Changes in gene expression were assayed by reverse transcription polymerase chain reaction (RT-PCR) at different passages in culture. Finally, differentiation was induced by addition of serum to the cultures and expression of markers for retinal cell types was detected by immunohistochemical staining with specific antibodies. Results: Dissociated cells from embryonic retina and CE at different postnatal ages generated primary nestin- and Pax6-immunoreactive neurosphere colonies in vitro in numbers that decreased with age. Embryonic and postnatal retina-derived RPCs and young CE-derived RSCs displayed self-renewal capacity, generating secondary neurosphere colonies. However, their self-renewal and proliferation capacity gradually decreased and they became more committed to differentiated states with subsequent passages. The expansion capacity of RPCs and RSCs was higher when they were maintained in monolayer culture. Porcine RPCs and RSCs could be induced to differentiate in vitro to express markers of retinal neurons and glia. Conclusions: Porcine retina and CE contain RPCs and RSCs which are undifferentiated, self-renewing and multipotent and which show characteristics similar to their human counterparts. Therefore, the pig could be a useful source of cells to further investigate the cell biology of RPCs and RSCs and it could be used as a non-primate large animal model for pre-clinical studies on stem cell-based approaches to regenerative medicine in the retina.

Ca2+-activated Cl- current in retinal arteriolar smooth muscle.

PURPOSE: To characterize the biophysical, pharmacologic, and functional properties of the Ca(2+)-activated Cl(-) current in retinal arteriolar myocytes. METHODS: Whole-cell perforated patch-clamp recordings were made from myocytes within intact isolated arteriolar segments. Arteriolar tone was assessed using pressure myography. RESULTS: Depolarizing of voltage steps to -40 mV and greater activated an L-type Ca(2+) current (I(Ca(L))) that was followed by a sustained current. Large tail currents (I(tail)) were observed on stepping back to -80 mV. The sustained current and I(tail) reversed close to 0 mV in symmetrical Cl(-) concentrations. The ion selectivity sequence for I(tail) was I(-)> Cl(-)> glucuronate. Outward I(tail) was sensitive to the Cl(-) channel blockers 9-anthracene-carboxylic acid (9-AC; 1 mM), 4-acetamido-4'-isothiocyanatostilbene-2,2'-disulfonic acid (SITS; 1 mM), and disodium 4,4'-diisothiocyanatostilbene-2,2'-disulfonate (DIDS; 1 mM), but only DIDS produced a substantial (78%) block of inward tail currents at -100 mV. I(tail) was decreased in magnitude when the normal bathing medium was substituted with Ca(2+)-free solution or if I(Ca(L)) was inhibited by 1 microM nimodipine. Caffeine (10 mM) produced large transient currents that reversed close to the Cl(-) equilibrium potential and were blocked by 1 mM DIDS or 100 microM tetracaine. DIDS had no effect on basal vascular tone in pressurized arterioles but dramatically reduced the level of vasoconstriction observed in the presence of 10 nM endothelin-1. CONCLUSIONS: Retinal arteriolar myocytes have I(Cl(Ca)), which may be activated by Ca(2+) entry through L-type Ca(2+) channels or Ca(2+) release from intracellular stores. This current appears to contribute to agonist-induced retinal vasoconstriction.

Segmentation of optic disc in retinal images using an improved gradient vector flow algorithm

Image segmentation plays an important role in the analysis of retinal images as the extraction of the optic disk provides important cues for accurate diagnosis of various retinopathic diseases. In recent years, gradient vector flow (GVF) based algorithms have been used successfully to successfully segment a variety of medical imagery. However, due to the compromise of internal and external energy forces within the resulting partial differential equations, these methods can lead to less accurate segmentation results in certain cases. In this paper, we propose the use of a new mean shift-based GVF segmentation algorithm that drives the internal/external energies towards the correct direction. The proposed method incorporates a mean shift operation within the standard GVF cost function to arrive at a more accurate segmentation. Experimental results on a large dataset of retinal images demonstrate that the presented method optimally detects the border of the optic disc.

Histological and ultrastructural investigation of retinal microaneurysm development in diabetic patients

BACKGROUND: Although microaneurysms are a clinicopathological hallmark of diabetic retinopathy, there have been few ultrastructural studies of these important lesions. As a result, knowledge of the mechanisms involved in the pathogenesis of microaneurysms remains fragmentary. This study provides histological and ultrastructural evidence of various stages in microaneurysm formation within the retinal vasculature. METHODS: The eyes of three type II diabetic patients, obtained within 24 hours of death, were studied by the trypsin digest technique. Eyes from two further type II diabetics were fixed in 2.5% glutaraldehyde within 12 hours of death and processed for electron microscopy. RESULTS: In the trypsin digest preparations, small saccular and fusiform microaneurysms were observed in the peripheral retinal. In the central retina, the microaneurysms ranged in morphology from thin walled, cellular forms to dense, acellular, hyalinised forms. Ultrastructurally, four distinct groups of microaneurysm were observed. Type I showed an extensive accumulation of polymorphonuclear cells into the lumen. The endothelium remained intact, although pericytes were invariably absent. Type II microaneurysms were typified by large numbers of red blood cells (RBCs) in the lumen. Endothelial cells and pericytes were completely absent. The type III microaneurysm was also non-perfused and contained aggregates of irregularly shaped RBC profiles and RBC breakdown products. Recanalisation by new vessels into the occluded lumen was observed in one microaneurysm. Type IV microaneurysms were almost or completely sclerosed, with extensive fibrosis and lipid infiltration into the lumen and basement membrane wall. CONCLUSION: This investigation describes several distinctive stages in the formation of microaneurysms during diabetic retinopathy. With reference to the pathogenesis of retinal microaneurysms, the interaction of various cell types is discussed and the significance of vascular cell death and localised hypertensive events highlighted.

Feedback via Ca²⁺-activated ion channels modulates endothelin 1 signaling in retinal arteriolar smooth muscle

PURPOSE: To investigate the role of feedback by Ca²?-sensitive plasma-membrane ion channels in endothelin 1 (Et1) signaling in vitro and in vivo. Methods. Et1 responses were imaged from Fluo-4-loaded smooth muscle in isolated segments of rat retinal arteriole using two-dimensional (2-D) confocal laser microscopy. Vasoconstrictor responses to intravitreal injections of Et1 were recorded in the absence and presence of appropriate ion channel blockers using fluorescein angiograms imaged using a confocal scanning laser ophthalmoscope. Results. Et1 (10 nM) increased both basal [Ca²?](i) and the amplitude and frequency of Ca²?-waves in retinal arterioles. The Ca²?-activated Cl?-channel blockers DIDS and 9-anthracene carboxylic acid (9AC) blocked Et1-induced increases in wave frequency, and 9AC also inhibited the increase in amplitude. Iberiotoxin, an inhibitor of large conductance (BK) Ca²?-activated K?-channels, increased wave amplitude in the presence of Et1 but had no effect on frequency. None of these drugs affected basal [Ca²?](i). The voltage-operated Ca²?-channel inhibitor nimodipine inhibited wave frequency and amplitude and also lowered basal [Ca²?](i) in the presence of Et1. Intravitreal injection of Et1 caused retinal arteriolar vasoconstriction. This was inhibited by DIDS but not by iberiotoxin or penitrem A, another BK-channel inhibitor. Conclusions. Et1 evokes increases in the frequency of arteriolar Ca²?-waves in vitro, resulting in vasoconstriction in vivo. These responses, initiated by release of stored Ca²?, also require positive feedback via Ca²?-activated Cl?-channels and L-type Ca²?-channels.

Fast Retinal Vessel Detection and Measurement Using Wavelets and Edge Location Refinement

The relationship between changes in retinal vessel morphology and the onset and progression of diseases such as diabetes, hypertension and retinopathy of prematurity (ROP) has been the subject of several large scale clinical studies. However, the difficulty of quantifying changes in retinal vessels in a sufficiently fast, accurate and repeatable manner has restricted the application of the insights gleaned from these studies to clinical practice. This paper presents a novel algorithm for the efficient detection and measurement of retinal vessels, which is general enough that it can be applied to both low and high resolution fundus photographs and fluorescein angiograms upon the adjustment of only a few intuitive parameters. Firstly, we describe the simple vessel segmentation strategy, formulated in the language of wavelets, that is used for fast vessel detection. When validated using a publicly available database of retinal images, this segmentation achieves a true positive rate of 70.27%, false positive rate of 2.83%, and accuracy score of 0.9371. Vessel edges are then more precisely localised using image profiles computed perpendicularly across a spline fit of each detected vessel centreline, so that both local and global changes in vessel diameter can be readily quantified. Using a second image database, we show that the diameters output by our algorithm display good agreement with the manual measurements made by three independent observers. We conclude that the improved speed and generality offered by our algorithm are achieved without sacrificing accuracy. The algorithm is implemented in MATLAB along with a graphical user interface, and we have made the source code freely available. 

An analysis of ABCR mutations in British patients with recessive retinal dystrophies

PURPOSE. Several reports have shown that mutations in the ABCR gene can lead to Stargardt disease (STGD)/fundus flavimaculatus (FFM), autosomal recessive retinitis pigmentosa (arRP), and autosomal recessive cone-rod dystrophy (arCRD). To assess the involvement of ABCR in these retinal dystrophies, the gene was screened in a panel of 70 patients of British origin. METHODS. Fifty-six patients exhibiting the STGD/FFM phenotype, 6 with arRP, and 8 with arCRD, were screened for mutations in the 50 exons of the ABCR gene by heteroduplex analysis and direct sequencing. Microsatellite marker haplotyping was used to determine ancestry. RESULTS. In the 70 patients analyzed, 31 sequence changes were identified, of which 20 were considered to be novel mutations, in a variety of phenotypes. An identical haplotype was associated with the same pair of in-cis alterations in 5 seemingly unrelated patients and their affected siblings with STGD/FFM. Four of the aforementioned patients were found to carry three alterations in the coding sequence of the ABCR gene, with two of them being in-cis. CONCLUSIONS. These results suggest that ABCR is a relatively polymorphic gene. Because putative mutations have been identified thus far only in 25 of 70 patients, of whom only 8 are compound heterozygotes, a large number of mutations have yet to be ascertained. The disease haplotype seen in the 5 patients carrying the same 'complex' allele is consistent with the presence of a common ancestor.