A generalization of the PAC learning in product probability spaces
نویسنده
چکیده
Three notions dependent theory, VC-dimension, and PAC-learnability have been found to be closely related. In addition to the known relation among these notions in model theory, finite combinatorics and probability theory, Chernikov, Palacin, and Takeuchi found a relation between n-dependence and VCn-dimension, which are generalizations of dependence and VC-dimension respectively. We are now working to find a generalization of PAC-learnability corresponding to the above two generalizations. This attempt is a joint work with Takayuki Kuriyama and Kota Takeuchi. In this article, we see basic definitions and known results as well as some examples.
منابع مشابه
Acceptable random variables in non-commutative probability spaces
Acceptable random variables are defined in noncommutative (quantum) probability spaces and some of probability inequalities for these classes are obtained. These results are a generalization of negatively orthant dependent random variables in probability theory. Furthermore, the obtained results can be used for random matrices.
متن کامل$C^{*}$-semi-inner product spaces
In this paper, we introduce a generalization of Hilbert $C^*$-modules which are pre-Finsler modules, namely, $C^{*}$-semi-inner product spaces. Some properties and results of such spaces are investigated, specially the orthogonality in these spaces will be considered. We then study bounded linear operators on $C^{*}$-semi-inner product spaces.
متن کاملOn Generalized Injective Spaces in Generalized Topologies
In this paper, we first present a new type of the concept of open sets by expressing some properties of arbitrary mappings on a power set. With the generalization of the closure spaces in categorical topology, we introduce the generalized topological spaces and the concept of generalized continuity and become familiar with weak and strong structures for generalized topological spaces. Then, int...
متن کاملAtomic Systems in 2-inner Product Spaces
In this paper, we introduce the concept of family of local atoms in a 2-inner product space and then this concept is generalized to an atomic system. Besides, a characterization of an atomic system lead to obtain a new frame. Actually this frame is a generalization of previous works.
متن کاملPredictive PAC Learning and Process Decompositions
We informally call a stochastic process learnable if it admits a generalization error approaching zero in probability for any concept class with finite VC-dimension (IID processes are the simplest example). A mixture of learnable processes need not be learnable itself, and certainly its generalization error need not decay at the same rate. In this paper, we argue that it is natural in predictiv...
متن کامل