litlas
Log inCreate account

Classification and Geometry of General Perceptual Manifolds

Authors: SueYeon Chung, Daniel D. Lee, Haim Sompolinsky - Physical Review X 2017 cited by 133

Perceptual manifolds arise when a neural population responds to an ensemble of sensory signals associated with different physical features (e.g., orientation, pose, scale, location, and intensity) of the same perceptual object. Object recognition and discrimination requires classifying the manifolds in a manner that is insensitive to variability within a manifold. How neuronal systems give rise to invariant object classification and recognition is a fundamental problem in brain theory as well as in machine learning. Here we study the ability of a readout network to classify objects from their perceptual manifold representations. We develop a statistical mechanical theory for the linear classification of manifolds with arbitrary geometry revealing a remarkable relation to the mathematics of conic decomposition. Novel geometrical measures of manifold radius and manifold dimension are introduced which can explain the classification capacity for manifolds of various geometries. The general theory is demonstrated on a number of representative manifolds, including L2 ellipsoids prototypical of strictly convex manifolds, L1 balls representing polytopes consisting of finite sample points, and orientation manifolds which arise from neurons tuned to respond to a continuous angle variable, such as object orientation. The effects of label sparsity on the classification capacity of manifolds are elucidated, revealing a scaling relation between label sparsity and manifold radius. Theoretical predictions are corroborated by numerical simulations using recently developed algorithms to compute maximum margin solutions for manifold dichotomies. Our theory and its extensions provide a powerful and rich framework for applying statistical mechanics of linear classification to data arising from neuronal responses to object stimuli, as well as to artificial deep networks trained for object recognition tasks.

Maximum Storage Capacityin Neural NetworksMaximum Storage Capacity in Neural NetworksThe space ofinteractions in neural…The space of interactions in neural network modelsProperties of neuralnetworks storing…Properties of neural networks storing spatially correlated patternsProceedings of IEEEConference on Computer…Proceedings of IEEE Conference on Computer Vision and Pattern RecognitionStatistical LearningTheoryStatistical Learning TheoryNonlinear DimensionalityReduction by Locally…Nonlinear Dimensionality Reduction by Locally Linear EmbeddingA Global GeometricFramework for Nonlinear…A Global Geometric Framework for Nonlinear Dimensionality ReductionUntangling invariantobject recognitionUntangling invariant object recognitionAdvances in NeuralInformation Processing…Advances in Neural Information Processing Systems 28Linear readout of objectmanifoldsLinear readout of object manifoldsOptimal Degrees ofSynaptic ConnectivityOptimal Degrees of Synaptic ConnectivityLearning Data Manifoldswith a Cutting Plane…Learning Data Manifolds with a Cutting Plane MethodLearning Data Manifoldswith a Cutting Plane…Learning Data Manifolds with a Cutting Plane MethodSeparability andGeometry of Object…Separability and Geometry of Object Manifolds in Deep Neural NetworksCounting the learnablefunctions of…Counting the learnable functions of geometrically structured dataStatistical learningtheory of structured…Statistical learning theory of structured dataModeling the Influenceof Data Structure on…Modeling the Influence of Data Structure on Learning in Neural Networks: The Hidden Manifold ModelBeyond the storagecapacity: data driven…Beyond the storage capacity: data driven satisfiability transitionEmergence of SeparableManifolds in Deep…Emergence of Separable Manifolds in Deep Language RepresentationsObject manifold geometryacross the mouse…Object manifold geometry across the mouse cortical visual hierarchySolvable Model for theLinear Separability of…Solvable Model for the Linear Separability of Structured DataLinear Classification ofNeural Manifolds with…Linear Classification of Neural Manifolds with Correlated VariabilityGeometry Linked toUntangling Efficiency…Geometry Linked to Untangling Efficiency Reveals Structure and Computation in Neural PopulationsRandom features andpolynomial rulesRandom features and polynomial rulesClassification andGeometry of General…Classification and Geometry of General Perceptual ManifoldsEarlier referencesFocus paperCiting papersOlderNewer

Click a node to pin it, click the empty canvas to go back to this paper, or hover to preview. Open a node’s page from its title.