Data Mining Practical Machine Learning Tools and Techniques

Social Network Analysis

Image classification is a complex process that may be affected by many factors. This paper examines current practices, problems, and prospects of image classification. The emphasis is placed on the summarization of major advanced classification approaches and the techniques used for improving classification accuracy. In addition, some important issues affecting classification performance are discussed. This literature review suggests that designing a suitable image-processing procedure is a prerequisite for a successful classification of remotely sensed data into a thematic map. Effective use of multiple features of remotely sensed data and the selection of a suitable classification method are especially significant for improving classification accuracy. Non-parametric classifiers such as neural network, decision tree classifier, and knowledge-based classification have increasingly become important approaches for multisource data classification. Integration of remote sensing, geographical information systems (GIS), and expert system emerges as a new research frontier. More research, however, is needed to identify and reduce uncertainties in the image-processing chain to improve classification accuracy.

/pdf/a-survey-of-image-classification-methods-and-techniques-for-5bc5ltr9dq.pdf

A survey of image classification methods and techniques for improving classification performance

Transactions of the Institute of British Geographers

The Visual Display of Quantitative Information

The landscape in which people live is made up of many features, which are named and have importance for cultural reasons. Promient among these are the naming of upland features such as mountains, but mountains are an enigmatic phenomenon which do not bear precise and repeatable definition. They have a vague spatial extent, and recent research has modelled such classes as spatial fuzzy sets. We take a specifically multiresolution approach to the definition of the fuzzy set membership of morphometric classes of landscape. We explore this idea with respect to the identification of culturally recognized landscape features of the English Lake District. Discussion focuses on peaks and passes, and the results show that the landscape elements identified in the analysis correspond to well-known landmarks included in a place name database for the area, although many more are found in the analysis than are named in the available database. Further analysis shows that a richer interrogation of the landscape can be achieved with Geographical Information Systems when using this method than using standard approaches. © Royal Geographical Society (with The Institute of British Geographers) 2004.

Where is Helvellyn? Fuzziness of multi‐scale landscape morphometry

We present a new technique for the visual exploration of origins (O) and destinations (D) arranged in geographic space. Previous attempts to map the flows between origins and destinations have suffered from problems of occlusion usually requiring some form of generalisation, such as aggregation or flow density estimation before they can be visualized. This can lead to loss of detail or the introduction of arbitrary artefacts in the visual representation. Here, we propose mapping OD vectors as cells rather than lines, comparable with the process of constructing OD matrices, but unlike the OD matrix, we preserve the spatial layout of all origin and destination locations by constructing a gridded two-level spatial treemap. The result is a set of spatially ordered small multiples upon which any arbitrary geographic data may be projected. Using a hash grid spatial data structure, we explore the characteristics of the technique through a software prototype that allows interactive query and visua...

/pdf/visualisation-of-origins-destinations-and-flows-with-od-maps-ght301181w.pdf

Visualisation of Origins, Destinations and Flows with OD Maps

Exploratory visual analysis is useful for the preliminary investigation of large structured, multifaceted spatio-temporal datasets. This process requires the selection and aggregation of records by time, space and attribute, the ability to transform data and the flexibility to apply appropriate visual encodings and interactions. We propose an approach inspired by geographical 'mashups' in which freely-available functionality and data are loosely but flexibly combined using de facto exchange standards. Our case study combines MySQL, PHP and the LandSerf GIS to allow Google Earth to be used for visual synthesis and interaction with encodings described in KML. This approach is applied to the exploration of a log of 1.42 million requests made of a mobile directory service. Novel combinations of interaction and visual encoding are developed including spatial 'tag clouds', 'tag maps', 'data dials' and multi-scale density surfaces. Four aspects of the approach are informally evaluated: the visual encodings employed, their success in the visual exploration of the dataset, the specific tools used and the 'mashup' approach. Preliminary findings will be beneficial to others considering using mashups for visualization. The specific techniques developed may be more widely applied to offer insights into the structure of multifarious spatio-temporal data of the type explored here. Extracting structure, meaning and insight from large, multifaceted, spatio-temporal datasets is a challenging task and such data are consequently underused in many application domains (1). This paper considers how the increasingly used 'mashup' approach to software and data integration can be applied to exploratory visualisation in order to address this underuse. In doing so, wider issues for the use of mashups in visualisation are identified. Exploratory visual analysis that proceeds in an iterative fashion can be a highly effective means of preliminary investigation (2,3). Such analyses require the means to select, inspect and modify visual encodings, with data transformed, filtered, sampled and aggregated according to time-, space- and attribute-based criteria. As the exploration proceeds, flexibility is needed to generate new samples, derivatives (such as alternate spatial aggregations), filterings and visual encodings in response to the insights and informal hypothesis gained and to integrate these with ancillary data. This process can be likened to Shneiderman's (4) "visual information-seeking mantra: overview first, zoom and filter, then details-on-demand". Off-the-shelf geographic information systems (GIS) provide some of the functionality, particularly for spatial processing, but less flexibility for geovisualization. The alternative of developing tools from scratch with the required flexibility is difficult, time-consuming and requires skills not possessed by many engaged in

https://openaccess.city.ac.uk/176/2/interactive visual exploration of a large spatio-temporal dataset.pdf

of a Large Spatio-Temporal Dataset: Reflections on a Geovisualization Mashup

Exploratory visual analysis is useful for the preliminary investigation of large structured, multifaceted spatio-temporal datasets. This process requires the selection and aggregation of records by time, space and attribute, the ability to transform data and the flexibility to apply appropriate visual encodings and interactions. We propose an approach inspired by geographical 'mashups' in which freely-available functionality and data are loosely but flexibly combined using de facto exchange standards. Our case study combines MySQL, PHP and the LandSerf GIS to allow Google Earth to be used for visual synthesis and interaction with encodings described in KML. This approach is applied to the exploration of a log of 1.42 million requests made of a mobile directory service. Novel combinations of interaction and visual encoding are developed including spatial 'tag clouds', 'tag maps', 'data dials' and multi-scale density surfaces. Four aspects of the approach are informally evaluated: the visual encodings employed, their success in the visual exploration of the dataset, the specific tools used and the 'mashup' approach. Preliminary findings will be beneficial to others considering using mashups for visualization. The specific techniques developed may be more widely applied to offer insights into the structure of multifarious spatio-temporal data of the type explored here.

/pdf/interactive-visual-exploration-of-a-large-spatio-temporal-4kcvr7u7jz.pdf

Interactive Visual Exploration of a Large Spatio-temporal Dataset: Reflections on a Geovisualization Mashup.

Existing treemap layout algorithms suffer to some extent from poor or inconsistent mappings between data order and visual ordering in their representation, reducing their cognitive plausibility. While attempts have been made to quantify this mismatch, and algorithms proposed to minimize inconsistency, solutions provided tend to concentrate on one-dimensional ordering. We propose extensions to the existing squarified layout algorithm that exploit the two-dimensional arrangement of treemap nodes more effectively. Our proposed spatial squarified layout algorithm provides a more consistent arrangement of nodes while maintaining low aspect ratios. It is suitable for the arrangement of data with a geographic component and can be used to create tessellated cartograms for geovisualization. Locational consistency is measured and visualized and a number of layout algorithms are compared. CIELab color space and displacement vector overlays are used to assess and emphasize the spatial layout of treemap nodes. A case study involving locations of tagged photographs in the Flickr database is described.

/pdf/spatially-ordered-treemaps-216ywk0h7w.pdf

Jo Wood

Papers

Where is Helvellyn? Fuzziness of multi‐scale landscape morphometry

Visualisation of Origins, Destinations and Flows with OD Maps

of a Large Spatio-Temporal Dataset: Reflections on a Geovisualization Mashup

Interactive Visual Exploration of a Large Spatio-temporal Dataset: Reflections on a Geovisualization Mashup.

Spatially Ordered Treemaps