Towards Manifold-Adaptive Learning
نویسندگان
چکیده
Inputs coming from high-dimensional spaces are common in many real-world problems such as a robot control with visual inputs. Yet learning in such cases is in general difficult, a fact often referred to as the “curse of dimensionality”. In particular, in regression or classification, in order to achieve a certain accuracy algorithms are known to require exponentially many samples in the dimension of the inputs in the worst-case [1]. The exponential dependence on the input dimension forces us to develop methods that are efficient in exploiting regularities of the data. Classically, smoothness is the best known example of such a regularity. In this abstract we outline two methods for two problems that are efficient in exploiting when the data points lie on a low dimensional submanifold of the input space.
منابع مشابه
A Curvature-based Manifold Learning Algorithm
Manifold learning aims to find a low dimensional parameterization for data sets which lie on nonlinear manifolds in a high-dimensional space. Applications of manifold learning include face recognition, image retrieval, machine learning, classification, visualization, and so on. By studying the existing manifold learning algorithms and geometric properties of local tangent space of a manifold, w...
متن کاملGeneralized Adaptive Dictionary Learning via Domain Shift Minimization
Visual data driven dictionaries have been successfully employed for various object recognition and classification tasks. However, the task becomes more challenging if the training and test data are from contrasting domains. In this paper, we propose a novel and generalized approach towards learning an adaptive and common dictionary for multiple domains. Precisely, we project the data from diffe...
متن کاملIntegration of exhaust manifold with engine cylinder head towards size and weight reduction
In this research, a new exhaust manifold and its cooling jackets is first designed for the integrated exhaust manifold into cylinder head (IEMCH) for a turbocharged engine. Then, the gas exchange and flow analysis is carried out numerically to evaluate the proper conditions for the exhaust gas and the coolant stream respectively. Finally, the entire engine parts are thermally analyzed to assure...
متن کاملA completed adaptive de-mixing algorithm on Stiefel manifold for ICA
Based on the one-bit-matching principle and by turning the de-mixing matrix into an orthogonal matrix via certain normalization, Ma et al proposed a one-bit-matching learning algorithm on the Stiefel manifold for independent component analysis [8]. But this algorithm is not adaptive. In this paper, an algorithm which can extract kurtosis and its sign of each independent source component directl...
متن کاملVariable Learning Rate Adaptive Sliding Mode Training Of Type-2 Fuzzy Neural Networks
This paper proposes a novel training method for the parameters of a type-2 fuzzy neural network (T2FNN) using sliding mode control theory with an adaptive learning rate. The implemented control structure consists of a conventional (PD) controller in parallel with a T2FNN. The former is responsible to guarantee global asymptotic stability in compact space and to form a sliding behavior. The outp...
متن کاملBoosting on Manifolds: Adaptive Regularization of Base Classifiers
In this paper we propose to combine two powerful ideas, boosting and manifold learning. On the one hand, we improve ADABOOST by incorporating knowledge on the structure of the data into base classifier design and selection. On the other hand, we use ADABOOST’s efficient learning mechanism to significantly improve supervised and semi-supervised algorithms proposed in the context of manifold lear...
متن کامل