Combinatorial generalization bounds
نویسندگان
چکیده
In this paper we propose a new combinatorial technique for obtaining data dependent generalization bounds. We introduce a splitting and connectivity graph (SC-graph) over the set of classifiers. In some cases the knowledge of this graph leads to an exact generalization bound. Typically, the knowledge of a little part of the SC-graph is sufficient for reasonable approximation of the bound. Being applied to a parametric set of conjunctive rules our bound helps to obtain more reliable classifiers as compositions of less overfitted rules.
منابع مشابه
Tight Combinatorial Generalization Bounds for Threshold Conjunction Rules
We propose a combinatorial technique for obtaining tight data dependent generalization bounds based on a splitting and connectivity graph (SC-graph) of the set of classifiers. We apply this approach to a parametric set of conjunctive rules and propose an algorithm for effective SC-bound computation. Experiments on 6 data sets from the UCI ML Repository show that SC-bound helps to learn more rel...
متن کاملNew combinatorial bounds for authenticationcodes and key predistribution
This paper provides new combinatorial bounds and characterizations of authentica-tion codes (A-codes) and key predistribution schemes (KPS). We rst prove a new lower bound on the number of keys in an A-code without secrecy, which can be thought of as a generalization of the classical Rao bound for orthogonal arrays. We also prove a new lower bound on the number of keys in a general A-code, whic...
متن کاملNew combinatorial bounds for authentication codesand key
This paper provides new combinatorial bounds and characterizations of authentication codes (A-codes) and key predistribution schemes (KPS). We rst prove a new lower bound on the number of keys in an A-code without secrecy, which can be thought of as a generalization of the classical Rao bound for orthogonal arrays. We also prove a new lower bound on the number of keys in a general A-code, which...
متن کاملSemi-Bandits with Knapsacks
We unify two prominent lines of work on multi-armed bandits: bandits with knapsacks and combinatorial semi-bandits. The former concerns limited “resources” consumed by the algorithm, e.g., limited supply in dynamic pricing. The latter allows a huge number of actions but assumes combinatorial structure and additional feedback to make the problem tractable. We define a common generalization, supp...
متن کاملPAC-Bayesian Bounds for Discrete Density Estimation and Co-clustering Analysis
We applied PAC-Bayesian framework to derive generalization bounds for co-clustering. The analysis yielded regularization terms that were absent in the preceding formulations of this task. The bounds suggested that co-clustering should optimize a trade-off between its empirical performance and the mutual information that the cluster variables preserve on row and column indices. Proper regulariza...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2011