Matches in SemOpenAlex for { <https://semopenalex.org/work/W30596594> ?p ?o ?g. }
Showing items 1 to 73 of
73
with 100 items per page.
- W30596594 abstract "For a given (possibly multivariate) random variable with a known (possibly multimodal) probability density function f(x), the 1 − α Highest Density Region (HDR) is the smallest possible region (or set of regions) which covers 100(1−α)% of the probability. Often, however, f(x) is unknown and only a sample of X (denoted by V ) is known. The algorithm that is discussed in this thesis aims to find an approximation to the 1−α HDR in these cases, and in a runtime which is polynomial in |V | (denoted by n). The algorithm makes use of the quantile approach, which states that given a sample y∗ of f(x), the d(1− α)ne elements of y∗ with the largest f -values are all in the 1− α HDR. The sample y∗ can be approximated by taking the reciprocal of w(v), the volume of the Voronoi cell of v, for all v ∈ V . The problem thus reduces to finding a subset X ⊂ V consisting of K connected components of the Delaunay triangulation graph of V (where K is the expected number of modes of f(x)) such that |X| = d(1− α)ne and such that the sum of w(v) for all v ∈ X is minimal. This problem is denoted as the connected component and cardinality-constrained Minimum Weight Connected Subgraph (CC-CC-MWCS) problem and it is an NP-hard problem. The CCCC-MWCS problem is solved exactly by two Dynamic Programming approaches and one Mixed-Integer Programming approach and it is solved approximately by a heuristic approach. The problem of finding the Delaunay triangulation graph of V and determining the volumes of the Voronoi cells of all v ∈ V can be done in polynomial time w.r.t. n. Both the runtime and memory usage of the DP approaches is exponential in n and they fail to solve the problem even for low values of n. The MIP approach can be formulated using a linear number of constraints and decision variables and is able to solve problem instances for larger values of n. The heuristic approach is polynomial in n and is shown to find solutions which are very close to optimal. Therefore, using the heuristic approach to solve the CC-CC-MWCS problem means that the full algorithm to find an HDR approximation becomes polynomial in n." @default.
- W30596594 created "2016-06-24" @default.
- W30596594 creator A5009399712 @default.
- W30596594 date "2014-09-19" @default.
- W30596594 modified "2023-09-26" @default.
- W30596594 title "An Algorithm for Approximating the Highest Density Region in d-Space" @default.
- W30596594 cites W1531327550 @default.
- W30596594 cites W1559408176 @default.
- W30596594 cites W1973109156 @default.
- W30596594 cites W1984406695 @default.
- W30596594 cites W1986698569 @default.
- W30596594 cites W2002055996 @default.
- W30596594 cites W2005243481 @default.
- W30596594 cites W2011235076 @default.
- W30596594 cites W201760388 @default.
- W30596594 cites W2069916913 @default.
- W30596594 cites W2072727857 @default.
- W30596594 cites W2114720692 @default.
- W30596594 cites W2153504150 @default.
- W30596594 cites W406610318 @default.
- W30596594 hasPublicationYear "2014" @default.
- W30596594 type Work @default.
- W30596594 sameAs 30596594 @default.
- W30596594 citedByCount "0" @default.
- W30596594 crossrefType "journal-article" @default.
- W30596594 hasAuthorship W30596594A5009399712 @default.
- W30596594 hasConcept C11413529 @default.
- W30596594 hasConcept C114614502 @default.
- W30596594 hasConcept C118615104 @default.
- W30596594 hasConcept C124101348 @default.
- W30596594 hasConcept C24881265 @default.
- W30596594 hasConcept C2524010 @default.
- W30596594 hasConcept C33923547 @default.
- W30596594 hasConcept C41008148 @default.
- W30596594 hasConcept C68010082 @default.
- W30596594 hasConcept C87117476 @default.
- W30596594 hasConceptScore W30596594C11413529 @default.
- W30596594 hasConceptScore W30596594C114614502 @default.
- W30596594 hasConceptScore W30596594C118615104 @default.
- W30596594 hasConceptScore W30596594C124101348 @default.
- W30596594 hasConceptScore W30596594C24881265 @default.
- W30596594 hasConceptScore W30596594C2524010 @default.
- W30596594 hasConceptScore W30596594C33923547 @default.
- W30596594 hasConceptScore W30596594C41008148 @default.
- W30596594 hasConceptScore W30596594C68010082 @default.
- W30596594 hasConceptScore W30596594C87117476 @default.
- W30596594 hasLocation W305965941 @default.
- W30596594 hasOpenAccess W30596594 @default.
- W30596594 hasPrimaryLocation W305965941 @default.
- W30596594 hasRelatedWork W1480687013 @default.
- W30596594 hasRelatedWork W1487551524 @default.
- W30596594 hasRelatedWork W155546234 @default.
- W30596594 hasRelatedWork W1589303579 @default.
- W30596594 hasRelatedWork W1596296817 @default.
- W30596594 hasRelatedWork W1607724231 @default.
- W30596594 hasRelatedWork W1808167818 @default.
- W30596594 hasRelatedWork W1885878332 @default.
- W30596594 hasRelatedWork W2001099600 @default.
- W30596594 hasRelatedWork W2002507856 @default.
- W30596594 hasRelatedWork W2021751389 @default.
- W30596594 hasRelatedWork W2026976813 @default.
- W30596594 hasRelatedWork W2176679876 @default.
- W30596594 hasRelatedWork W2336610657 @default.
- W30596594 hasRelatedWork W2726267431 @default.
- W30596594 hasRelatedWork W2949859737 @default.
- W30596594 hasRelatedWork W2951636664 @default.
- W30596594 hasRelatedWork W2951826107 @default.
- W30596594 hasRelatedWork W3128304991 @default.
- W30596594 hasRelatedWork W797882439 @default.
- W30596594 isParatext "false" @default.
- W30596594 isRetracted "false" @default.
- W30596594 magId "30596594" @default.
- W30596594 workType "article" @default.