Robust Distance-Based Clustering with Applications to Spatial Data Mining

Estivill‐Castro, Vladimir; Houle, Michael E.

doi:10.1007/s00453-001-0010-1

Cited by 32 publications

(39 citation statements)

References 70 publications

(87 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, the method suffers from severe limitations when clustering large spatial dataset [5] due to the complexity of computing distance between medoid points representing each pair of clusters. These efficiency drawbacks are partially alleviated when adopting both proximity and density information to achieve high quality spatial clusters in a sub-quadratic time without requiring the user to a-priori specify the number of clusters [7]. Similarly, DBSCAN [6] exploits density information to efficiently detect clusters of arbitrary shape from point spatial data with noise.…”

Section: Background and Motivationmentioning

confidence: 99%

See 1 more Smart Citation

Spatial Clustering of Structured Objects

Malerba

Appice

Varlaro

et al. 2005

Inductive Logic Programming

View full text Add to dashboard Cite

Abstract. Clustering is a fundamental task in Spatial Data Mining where data consists of observations for a site (e.g. areal units) descriptive of one or more (spatial) primary units, possibly of different type, collected within the same site boundary. The goal is to group structured objects, i.e. data collected at different sites, such that data inside each cluster models the continuity of socio-economic or geographic environment, while separate clusters model variation over the space. Continuity is evaluated according to the spatial organization arising in data, namely discrete spatial structure, expressing the (spatial) relations between separate sites implicitly defined by their geometrical representation and positioning. Data collected within sites that are (transitively) connected in the discrete spatial structure are clustered together according to the similarity on multi-relational descriptions representing their internal structure. CORSO is a novel spatial data mining method that resorts to a multi-relational approach to learn relational spatial data and exploits the concept of neighborhood to capture relational constraints embedded in the discrete spatial structure. Relational data are expressed in a firstorder formalism and similarity among structured objects is computed as degree of matching with respect to a common generalization. The application to real-world spatial data is reported.

show abstract

Section: Background and Motivationmentioning

confidence: 99%

“…length(X 2 ) = [7,10] which differ only in the value of a single selector (length), the first-order clause obtained by generalizing the pairs of comparable selectors in both H 1 and H 2 is: Example 3: Let us consider T C that is the set of first-order clauses including:…”

Section: Examplementioning

confidence: 99%

Spatial Clustering of Structured Objects

Malerba

Appice

Varlaro

et al. 2005

Inductive Logic Programming

View full text Add to dashboard Cite

show abstract

“…Optimization in the K-means sense as commonly used 10,12,13 would require that the weighted mean center of all straw assigned to a specifi c plant site be identical to the SSAO site was that the variance in straw assignment per site was 55%…”

Section: Validation Of Ssao and Comparison With K-meansmentioning

confidence: 99%

“…In the fi rst case, spatial clustering algorithms such as the relatively simple K-means method 10,11 and its more elaborate modifi cations 12,13 produce facility siting solutions that are locally optimized and oft en close to globally optimized, 14 within constraints imposed by practical limitations on computational time. Th e general goal of such optimizations is maximization of the area covered by service facilities within prescribed limits on response time 15 or minimization of the total cost of providing service.…”

mentioning

confidence: 99%

Geospatial identification of optimal straw‐to‐energy conversion sites in the Pacific Northwest

Mueller‐Warrant¹,

Banowetz²,

Whittaker³

2010

Biofuels Bioprod Bioref

View full text Add to dashboard Cite

Previous attempts to develop straw-based bioenergy systems have stumbled at costs of transporting this low-density resource to large-scale, centralized facilities. Success in developing small-scale, distributed technologies (e.g. syngas or pyrolysis bio-oil) that reduce these costs will depend on closely matching system requirements to spatial distribution of available straw. We analyzed straw distribution in the Pacifi c Northwest to identify optimal sites for facilities ranging from a pilot plant currently under development to larger ones of previous studies. Sites for plants with capacities of 1, 10, or 100 million kg straw y -1 were identifi ed using a 'lowest-hanging-fruit' iterative siting process in which the location of maximum density of straw over an appropriately sized neighborhood was identifi ed, distance from that point necessary to include desired quantity of straw measured, straw assigned to that plant removed from the raster, and the process repeated until all available straw had been assigned. Compared to K-means, our new method sited the fi rst 44% of plants at superior locations in terms of local straw density (i.e. lower transportation costs) and the next 39% at equivalent locations. K-means produced better locations for the fi nal 17% of plants along with superior average results. For the smallest facilities at locations defi ned by 3-year average available straw density, 1.2 km buffers were adequate to provide straw for the fi rst 10% of plants, with twice that distance suffi cient for the fi rst 70%. For the largest plants, 12 km buffers satisfi ed the fi rst 10% of plants, with 24 km buffers satisfying the fi rst 60%. Buffer distances exceeded 68 km for the fi nal 20% of the largest plants. Siting patterns for the smallest plants were more evenly distributed than larger ones, suggesting that farm-scale technology may be more politically appealing. Smaller plants, however, suffered from higher year-to-year variability in straw supply within pre-defi ned distances. Published in

show abstract

“…OUTPUT: the detected preference or that no preference was found Complexity: If n denotes the number of tuples in the log relation and k is the number of different values, the k-means clustering needs O(k 2 ) [2] leading to the overall complexity O(n + k 2 ). Typically, we have k << n and with it the complexity O(n).…”

Section: Algorithm 1: Miner For Categorical Preferences In Static Dommentioning

confidence: 99%

Preference Mining: A Novel Approach on Mining User Preferences for Personalized Applications

Holland

Ester

Kießling

2003

Knowledge Discovery in Databases: PKDD 2003

View full text Add to dashboard Cite

Abstract.Advanced personalized e-applications require comprehensive knowledge about their user's likes and dislikes in order to provide individual product recommendations, personal customer advice and custom-tailored product offers. In our approach we model such preferences as strict partial orders with "A is better than B" semantics, which has been proven to be very suitable in various e-applications. In this paper we present novel Preference Mining techniques for detecting strict partial order preferences in user log data. The main advantage of our approach is the semantic expressiveness of the Preference Mining results. Experimental evaluations prove the effectiveness and efficiency of our algorithms. Since the Preference Mining implementation uses sophisticated SQL statements to execute all data-intensive operations on database layer, our algorithms scale well even for large log data sets. With our approach personalized e-applications can gain valuable knowledge about their customers' preferences, which is essential for a qualified customer service.

show abstract

Robust Distance-Based Clustering with Applications to Spatial Data Mining

Cited by 32 publications

References 70 publications

Spatial Clustering of Structured Objects

Spatial Clustering of Structured Objects

Geospatial identification of optimal straw‐to‐energy conversion sites in the Pacific Northwest

Preference Mining: A Novel Approach on Mining User Preferences for Personalized Applications

Contact Info

Product

Resources

About