Showing papers by "École Normale Supérieure published in 1995"

Comparison of learning algorithms for handwritten digit recognition

Abstract: COMPARISON OF LEARNINGALGORITHMS FOR HANDWRITTEN DIGITRECOGNITIONY. LeCun, L. Jackel, L. Bottou, A. Brunot, C. Cortes,J. Denker, H. Drucker, I. Guyon, U. M uller,E. Sackinger, P. Simard, and V. VapnikBell Lab oratories, Holmdel, NJ 07733, USAEmail: yann@research.att.comAbstractThis pap er compares the p erformance of several classi er algorithmson a standard database of handwritten digits. We consider not only rawaccuracy, but also rejection, training time, recognition time, and memoryrequirements.1

Coded orthogonal frequency division multiplex [TV broadcasting]

Abstract: Technological evolution and the ever-increasing demand for higher-quality services give broadcasters a strong incentive to completely digitize their broadcasting networks. This digitization, which is already well advanced in many program production areas and transmission links, now has to be extended to complete the last link in the broadcast chain; i.e., from broadcast transmitter to consumer receivers. It is therefore necessary to develop wholly new techniques for the broadcasting of digitally coded TV programmes. Thus an efficient baseband digital coding must be combined with a robust digital modulation and channel coding scheme that can meet the requirements of every mode of broadcast reception. This article presents the research work related to the coded orthogonal frequency division multiplex (COFDM) technology, which has now been completed in the field of digital radio (DAB), and which is under progress in the field of digital terrestrial TV. >

Opening up Decision Making: The View from the Black Stool

Abstract: Set on its current course thirty years ago by Herbert Simon's notions of bounded rationality and sequential stages, the research literature of organizational decision making is claimed in this paper to have suffered from three major limitations labeled reification, dehumanization, and isolation. In particular, it has been stuck along a continuum between the cerebral rationality of the stage theories at one end and the apparent irrationality of the theory of organized anarchies at the other. This paper seeks to open up decision making in three respects. First, the concept of “decision” is opened up to the ambiguities that surround the relationship between commitment and action. Second, the decision maker is opened up to history and experience, to affect and inspiration, and especially to the critical role of insight in transcending the bounds of cerebral rationality. Third, the process of decision making is opened up to a host of dynamic linkages, so that isolated traces of single decisions come to be seen as interwoven networks of issues. The paper concludes with a plea to open up research itself to the development of richer theory on these important processes.

Learning algorithms for classification: A comparison on handwritten digit recognition

Abstract: This paper compares the performance of several classi er algorithms on a standard database of handwritten digits. We consider not only raw accuracy, but also training time, recognition time, and memory requirements. When available, we report measurements of the fraction of patterns that must be rejected so that the remaining patterns have misclassi cation rates less than a given threshold.

Stretching DNA with a receding meniscus: Experiments and models.

Abstract: A detailed experimental and theoretical analysis of the alignment of grafted DNA molecules by a moving meniscus is presented. The existence and extent of the stretching (up to 2.14 times the unstretched length) depends critically on the properties of the surface. Molecules grafted at both ends exhibit a looplike shape which is scale invariant. An elastic model of this process, which we have called molecular combing, is introduced which (a) yields the extension force on various surfaces, ( b) yields a value for the tensile strength of DNA, 476 6 84 pN, and (c) describes the shape of the loops with no fitting parameters.

The social context of reasoning : conversational inference and rational judgment

Abstract: Most psychologists conceive of judgment and reasoning as cognitive processes, which go on “in the head” and involve intrapsychic information processing (e.g., Kahneman, Slovic, and Tversky, 1982; Nisbett and Ross, 1980). Although it is incontestable that processes of attention, memory, and inference underpin judgment and reasoning, psychologists have perhaps overlooked the extent to which these mental processes are governed by higher-level assumptions about the social context of the information to be processed. On the other hand, philosophers have in recent years drawn attention to the extent to which reasoning from ordinary language is shaped by the nature of social interaction and conversation (Austin, 1962; Grice, 1975; Hart and Honore, 1959/1985; Mackie, 1974; Searle, 1969; Strawson, 1952). These higher-level assumptions can determine what we attend to, which memories we search, and what kinds of inference we draw. Consider the way the word family can be differentially interpreted according to context and thus lead to seemingly inconsistent judgments expressed in a conversational exchange (cf. Strack, Martin, and Schwarz, 1988): Q. How is your family? A. Fairly well, thank you. A married man might reply this way if he considers that his wife has recently been saddened by the loss of a close friend but that his two children are in good form. The respondent interprets family to mean the wife and kids .

The sulfate‐CCN‐cloud albedo effect

Abstract: Aerosol particles, such as sulfate aerosols, can act as cloud condensation nuclei (CCN). The CCN spectrum and the water vapour supply in a cloud determine the cloud droplet number concentration (CDNC) and hence the shortwave optical properties of low-level liquid clouds. The capability of anthropogenic aerosols to increase cloud reflectivity and thereby cool the Earth's surface is referred to as the indirect effect of anthropogenic aerosols. To obtain an estimate of this effect on climate, we empirically relate the CDNC, and thus the cloud optical properties, of two general circulation models (GCM) to the sulfate aerosol mass concentration derived from a chemical transport model. Based on a series of model experiments, the normalized globally averaged indirect forcing is about − 1 W/m 2 and ranges from – 0.5 to − 1.5 W/m 2 in both GCMs for different experiments. However, it is argued that the total uncertainty of the forcing is certainly larger than this range. The overall agreement between the two climate models is good, although the geographical distributions of the forcing are somewhat different. The highest forcings occur in and off the coasts of the polluted regions of the Northern Hemisphere. The regional distribution of the forcing and the land/sea contrast are very sensitive to the choice of the CDNC-sulfate mass relationship. The general patterns of the forcing, and the appropriateness of the different CDNC-sulfate mass relationships, are assessed. We also examine the simulated droplet effective radii and compare them with satellite retrievals. DOI: 10.1034/j.1600-0889.47.issue3.1.x

Elastic theory of flux lattices in the presence of weak disorder.

Abstract: The effect of weak impurity disorder on flux lattices at equilibrium is studied quantitatively in the absence of free dislocations using both the Gaussian variational method and the renormalization group. Our results for the mean-square relative displacements B\ifmmode \tilde{}\else \~{}\fi{}(x)=〈u(x)-u(0)${\mathrm{〉}}^{2}$\ifmmode\bar\else\textasciimacron\fi{} clarify the nature of the crossovers with distance. We find three regimes: (i) a short distance regime (``Larkin regime'') whre elasticity holds, (ii) an intermediate reigme (``random manifold'') where vortices are pinned independently, and (iii) a large distance, quasiordered regime where the periodicity of the lattice becomes important. In the last regime we find universal logarithmic growth of displacements for 2d4: B\ifmmode \tilde{}\else \~{}\fi{}(x)\ensuremath{\sim}${\mathit{A}}_{\mathit{d}}$ln\ensuremath{\Vert}x\ensuremath{\Vert} and persistence of algebraic quasi-long-range translational order. The functional renormalization group to O(\ensuremath{\epsilon}=4-d) and the variational method, when they can be compared, agree within 10% on the value of ${\mathit{A}}_{\mathit{d}}$. In d=3 we compute the function describing the crossover between the three reigmes. We discuss the observable signature of this crossover in decoration experiments and in neutron-diffraction experiments on flux lattices. Qualitative arguments are given suggesting the existence for weak disorder in d=3 of a ``Bragg glass'' phase without free dislocations and with algebraically divergent Bragg peaks. In d=1+1 both the variational method and the Cardy-Ostlund renormalization group predict a glassy state below the same transition temperature T=${\mathit{T}}_{\mathit{c}}$, but with different B\ifmmode \tilde{}\else \~{}\fi{}(x) behaviors. Applications to d=2+0 systems and experiments on magnetic bubbles are discussed.

Introduction to Geochemical Modeling

Abstract: Foreword Preface Introduction 1. Mass balance, mixing and fractionation 2. Linear algebra 3. Useful numerical analysis 4. Probability and statistics 5. Inverse methods 6. Modeling chemical equilibrium 7. Dynamic systems 8. Transport, advection and diffusion 9. Trace elements in magmatic processes References Index.

Recent advances in molecular recognition and signal transduction of active peptides: Receptors for opioid peptides

Abstract: 1. Opioid peptides are a family of structurally related neuromodulators which play a major role in the control of nociceptive pathways. These peptides act through membrane receptors of the nervous system, defined as mu, delta and kappa and endowed with overlapping but distinct pharmacological, anatomical and functional properties. 2. Recent cloning of an opioid receptor gene family has opened the way to the use of recombinant DNA technology at the receptor level. 3. This review focuses on the molecular cloning and functional characterization of opioid receptors and provides first insights into molecular aspects of opioid peptide recognition and signal transduction mechanisms, using the cloned receptors as investigation tools.

From 3D magnetic resonance images to structural representations of the cortex topography using topology preserving deformations

Abstract: We propose an algorithm allowing the construction of a structural representation of the cortical topography from a T1-weighted 3D MR image. This representation is an attributed relational graph (ARG) inferred from the 3D skeleton of the object made up of the union of gray matter and cerebro-spinal fluid enclosed in the brain hull. In order to increase the robustness of the skeletonization, topological and regularization constraints are included in the segmentation process using an original method: the homotopically deformable regions. This method is halfway between deformable contour and Markovian segmentation approaches. The 3D skeleton is segmented in simple surfaces (SSs) constituting the ARG nodes (mainly cortical folds). The ARG relations are of two types: first, theSS pairs connected in the skeleton; second, theSS pairs delimiting a gyrus. The described algorithm has been developed in the frame of a project aiming at the automatic detection and recognition of the main cortical sulci. Indeed, the ARG is a synthetic representation of all the information required by the sulcus identification. This project will contribute to the development of new methodologies for human brain functional mapping and neurosurgery operation planning.

Thin crust, ultramafic exposures, and rugged faulting patterns at the Mid-Atlantic Ridge (22°–24°N)

Abstract: Off-axis rock sampling in the lat 22°–24° N region of the Mid-Atlantic Ridge shows that the emplacement of mantle-derived rocks in the sea floor has been a common process there for the past few million years. We find a good correlation between domains of positive residual gravity anomalies (inferred to have a thin crust) and the distribution of ultramafic samples. We also find that thin-crust domains have a rugged topography, thought to reflect strong tectonic disruption. We propose that these thin-crust domains are made of tectonically uplifted ultramafic rocks, with gabbroic intrusions and a thin basaltic cover. We also suggest that strong tectonic disruption may be a direct consequence of the lithological and rheological heterogeneity of these thin-crust domains.

Mapping of a gene for long QT syndrome to chromosome 4q25-27.

Abstract: Long QT syndrome (LQTS) is a heterogeneous inherited disorder causing syncope and sudden death from ventricular arrhythmias. A first locus for this disorder was mapped to chromosome 11p15.5. However, locus heterogeneity has been demonstrated in several families, and two other loci have recently been located on chromosomes 7q35-36 and 3p21-24. We used linkage analysis to map the locus in a 65-member family in which LQTS was associated with more marked sinus bradycardia than usual, leading to sinus node dysfunction. Linkage to chromosome 11p15.5, 7q35-36, or 3p21-24 was excluded. Positive linkage was obtained for markers located on chromosome 4q25-27. A maximal LOD score of 7.05 was found for marker D4S402. The identification of a fourth locus for LQTS confirms its genetic heterogeneity. Locus 4q25-27 is associated with a peculiar phenotype within the LQTS entity.

Splitting of high-Q Mie modes induced by light backscattering in silica microspheres

Abstract: We have observed that very high-Q Mie resonances in silica microspheres are split into doublets. This splitting is attributed to internal backscattering that couples the two degenerate whispering-gallery modes propagating in opposite directions along the sphere equator. We have studied this doublet structure by high-resolution spectroscopy. Time-decay measurements have also been performed and show a beat note corresponding to the coupling rate between the clockwise and counterclockwise modes. A simple model of coupled oscillators describes our data well, and the backscattering efficiency that we measure is consistent with what is observed in optical fibers.

Archaean subduction inferred from seismic images of a mantle suture in the Superior Province

Abstract: PLATE tectonics provides the basis for the interpretation of most current terrestrial tectonic activity, and is widely accepted as having been active over much of the Earth's history1. Yet the timing of initiation of this process is subject to debate2–9. So far, the earliest seismic evidence for plate tectonics has come from a fossil mantle suture in the Svecofennian orogen (1.89 Gyr ago)10 and from inferred plate convergence, subduction and accretion in the Trans-Hudson orogen (1.91–1.79 Gyr ago)11. As yet, seismic data from Archaean areas have been able to demonstrate only the importance of compression in the construction of the continental crust12–15. Here we present seismic data from a collision zone in the Superior Province of Canada, involving the Abitibi granite–greenstone Subprovince and the plutonic, arc-related Opatica belt. We interpret dipping seismic reflections that extend 30km into the mantle as representing a relict 2.69-Gyr-old suture associated with subduction. Although crustal structure, lithospheric thicknesses and convergence rates may have differed from those seen today, these seismic data provide direct evidence that plate tectonics was active in late Archaean times, 800 Myr earlier than indicated by previous seismic reflection surveys.

Optimality conditions for bilevel programming problems

Abstract: The bilevel programming problem (BLPP) is a sequence of two optimization problems where the constraint region of the upper level problem is determined implicitly by the solution set to the lower level problem. To obtain optimality conditions, we reformulate BLPP as a single level mathematical programming problem (SLPP) which involves the value function of the lower level problem. For this mathematical programming problem, it is shown that in general the usual constraint qualifications do not hold and the right constraint qualification is the calmness condition. It is also shown that the linear bilevel programming problem and the minmax problem satisfy the calmness condition automatically. A sufficient condition for the calmness for the bilevel programming problem with quadratic lower level problem and nondegenerate linear complementar¬ity lower level problem are given. First order necessary optimality condition are given using nonsmooth analysis. Second order sufficient optimality conditions are also give...

Chapter 2 Time constrained routing and scheduling

Abstract: Publisher Summary Time constrained routing and scheduling problems are encountered in a variety of industrial and service sector applications, ranging from logistics and transportation systems to material handling systems in manufacturing. The traveling salesman problem with time windows has applications in single and multiple vehicle problems. The vehicle routing problem with time windows has many industrial applications including those where dock availability is a bottleneck such as for distribution centers. This chapter describes the significant advances made in time constrained routing and scheduling. In terms of solution methodology capable of solving realistic size problems, this field has seen a natural progression from ad-hoc methods to simple heuristics, to optimization-based heuristics and recently optimal algorithms. The chapter discusses fixed schedule problems and develops in detail the Dantzig-Wolfe decomposition/column generation approach which will then be applied to many of the other problem types. The vehicle routing problem with time windows and several important problem variants including the multiple traveling salesman problems is explored. The chapter examines a unified framework for fleet and crew scheduling problems.

Cooperative molecular motors.

Abstract: We present a simple stochastical model for motor molecules that cooperate in large groups. This model could apply for actin-myosin motors in muscles and for motility assays with a high concentration of motor molecules. We calculate the dependence of the velocity on the applied force as a function of ATP concentration and show the existence of a dynamical phase transition allowing for spontaneous directed motion even if the system is spatially symmetric. In the symmetric case, the problem is isomorphous to a paramagnet-ferromagnet transition, in the asymmetric case to a liquid-vapor transition.