Non-hierarchical Technique Research Articles

Identificação do padrão de comportamento de moradores em dados reais para simulações de desempenho térmico: estudo de caso em habitações de interesse social

Abstract The way residents occupy and operate houses influences the energy consumption. The objective of this paper is to findpatterns of occupant behaviour in actual data for thermal performance simulation.Data on occupancy of rooms and operation of doors and windows were obtained through a database created by means of application of questionnaires in low-income houses located in Florianópolis, southern Brazil. The reference profiles were obtained using cluster analysis, hierarchical and non-hierarchical techniques combined. Such profiles were submitted to computer simulations. The results showed significant variability within the clusters regarding the occupancy and operation of doors and windows. It was possible to verify the impact that different profiles have on the performance of the house, either due to occupancy or heat losses and gains from air changes through doors and windows. The combination of these effects resulted in some profiles that were highly vulnerable to external temperature conditions, while others were able to maintain the internal temperatures more constant. It was possible to verify that the use of reference profiles based on actual data lead to more reliable performance indicators.

Cluster Analysis for Data Processing in Educational Research

Quantitative approaches to educational research have been undervalued and consequently less widely used. In this sense, this paper aims to present and analyze the techniques of Cluster Analysis as a possibility for research in sciences area. Therefore, the main hierarchical and non-hierarchical techniques of Cluster Analysis are presented, as well as some of their applications in educational research found in the literature. Cluster Analysis is adequate to simplify or elaborate hypotheses on massive data, such as large-scale educational research. The studies in the area of education that used Cluster Analysis methods proved to be fruitful to elicit results that collaborate with the area.

Regional variation in household income

The aim of the paper is to determine the scale of intervoivodship variation in terms of the level of household income, taking into account the sources of this income. To evaluate regional differences related to the economic situation of households, the authors used data on the sources and level of income per person in a household in a voivodship. Average values of income for each identified source were calculated on the basis of microdata from the 2016 Household Budget Survey carried out by Statistics Poland. For the purpose of the study, the authors employed k-medoid method, one of non-hierarchical techniques of division. It was demonstrated that of all the diagnostic features adopted in the study, it was the level of income from permanent employment abroad which differentiated among particular clusters of voivodships to the largest extent. The average value for this feature in cluster 2 was PLN 5000,70 per person in the household. The performed grouping of voivodships showed that the extent to which households search for additional sources of income depends on the economic potential of the region and available opportunities.

Method for obtaining reference buildings

Reference buildings are useful tools for evaluating energy saving measures of an entire building stock. The objective of this study is to develop a method for obtaining reference buildings for the low-income housing stock in Florianópolis, Southern Brazil. Field data collection was performed in order to build a database on geometrical features of houses. Reference buildings were obtained using cluster analysis, which aims to find subgroups within a sample. For that, hierarchical and non-hierarchical techniques were applied in combination. Reference buildings were taken as the nearest houses to the mean of their clusters. Computer simulation using Energyplus were performed to verify the accuracy of the found reference buildings. The 120 housing units sample database obtained from the survey resulted into two reference buildings after cluster analysis: a 76m2 house, with living room, kitchen and three bedrooms, and a 37m2 house, with combined living room and kitchen and two bedrooms. Simulations have shown that the reference buildings can represent their cluster properly, since the degree-hour values obtained for them were similar to the housing median sample. Cluster analysis proved to be a technique useful for obtaining reference buildings, although one should be very careful in selecting the variables involved in the analysis.

Discovering hierarchical object models from captioned images

We address the problem of automatically learning the recurring associations between the visual structures in images and the words in their associated captions, yielding a set of named object models that can be used for subsequent image annotation. In previous work, we used language to drive the perceptual grouping of local features into configurations that capture small parts (patches) of an object. However, model scope was poor, leading to poor object localization during detection (annotation), and ambiguity was high when part detections were weak. We extend and significantly revise our previous framework by using language to drive the perceptual grouping of parts, each a configuration in the previous framework, into hierarchical configurations that offer greater spatial extent and flexibility. The resulting hierarchical multipart models remain scale, translation and rotation invariant, but are more reliable detectors and provide better localization. Moreover, unlike typical frameworks for learning object models, our approach requires no bounding boxes around the objects to be learned, can handle heavily cluttered training scenes, and is robust in the face of noisy captions, i.e., where objects in an image may not be named in the caption, and objects named in the caption may not appear in the image. We demonstrate improved precision and recall in annotation over the non-hierarchical technique and also show extended spatial coverage of detected objects.

On The Application of Fuzzy Clustering for Crime Hot Spot Detection

One of the fundamental challenges in crime mapping and analysis is pattern recognition. Efforts and methods to detect crime hot-spots, or geographic areas of elevated criminal activity, are wide ranging. For aggregate data, such as total crime events in a census tract(s), measures of spatial autocorrelation have proven useful. For disaggregate data (i.e. individual crime events), kernel density smoothing and non-hierarchical cluster analysis (e.g. k-means), are widely used. Non-hierarchical techniques are particularly effective in delineating geographic space into areas of higher or lower crime concentrations, because each observation is assigned to one and only one cluster. The resulting set of partitions provides clear-cut spatial boundaries that can be used for hot-spot evaluation and interpretation. However, the strength of non-hierarchical methods can also be viewed as a weakness. Although the hard-clustering of observations into a set of discrete clusters is helpful, there are many cases where ambiguity exists in the data. In such cases, a more generalized approach for hot-spot detection would be helpful. The purpose of this paper is to explore the use of a generalized partitioning method known as fuzzy clustering for hot-spot detection. Functional and visual comparisons of fuzzy clustering and two hard-clustering approaches (medoid and k-means), across a range of cluster values are analyzed. The empirical results suggest that a fuzzy clustering approach is better equipped to handle intermediate cases and spatial outliers.

PAQ: Partition Analysis of Quasispecies.

The complexities of genetic data may not be accurately described by any single analytical tool. Phylogenetic analysis is often used to study the genetic relationship among different sequences. Evolutionary models and assumptions are invoked to reconstruct trees that describe the phylogenetic relationship among sequences. Genetic databases are rapidly accumulating large amounts of sequences. Newly acquired sequences, which have not yet been characterized, may require preliminary genetic exploration in order to build models describing the evolutionary relationship among sequences. There are clustering techniques that rely less on models of evolution, and thus may provide nice exploratory tools for identifying genetic similarities. Some of the more commonly used clustering methods perform better when data can be grouped into mutually exclusive groups. Genetic data from viral quasispecies, which consist of closely related variants that differ by small changes, however, may best be partitioned by overlapping groups. We have developed an intuitive exploratory program, Partition Analysis of Quasispecies (PAQ), which utilizes a non-hierarchical technique to partition sequences that are genetically similar. PAQ was used to analyze a data set of human immunodeficiency virus type 1 (HIV-1) envelope sequences isolated from different regions of the brain and another data set consisting of the equine infectious anemia virus (EIAV) regulatory gene rev. Analysis of the HIV-1 data set by PAQ was consistent with phylogenetic analysis of the same data, and the EIAV rev variants were partitioned into two overlapping groups. PAQ provides an additional tool which can be used to glean information from genetic data and can be used in conjunction with other tools to study genetic similarities and genetic evolution of viral quasispecies.

Using cluster analysis to develop a healing typology in vascular ulcers

A clinically meaningful typology of healing was developed to assist nurses in assessing patients' healing potential. A descriptive correlational design was used to determine similarities in healing characteristics of older people with venous, venous-arterial, and arterial leg ulcers (n = 156). Hierarchical clustering techniques dendrogram using Ward's method suggested that three clusters existed based on the ankle/brachial pressure index, liposclerosis (hardening and induration of the skin of the lower limb), edema, and wound characteristics; thus a healing typology was potentially formed. Nonhierarchical techniques such as analysis of variance examined cluster differences for the initial ulcer area (p < 0.013) and suggested that these groupings were reflective of differing ulcer areas. However, significant differences in the rate of healing (difference between ulcer surface area in week 4 and week 1) (measured by topographic mapping) group comparisons were not significant at the 0.05 level. Clinical differences in the healing rate were apparent, and three individual profiles of leg ulcer sufferers based on the three differing clusters of factors were outlined: rapid healers, slow healers, and nonhealers. Further research into the clustering of these factors and their application to longer healing times or time to complete healing may be significant and support the utility of this typology.

The analysis of large scale data taken from the world groundnut (Arachis hypogaea L.) germplasm collection I. Two-way quantitative data

Data associated with germplasm collections are typically large and multivariate with a considerable number of descriptors measured on each of many accessions. Pattern analysis methods of clustering and ordination have been identified as techniques for statistically evaluating the available diversity in germplasm data. While used in many studies, the approaches have not dealt explicitly with the computational consequences of large data sets (i.e. greater than 5000 accessions). To consider the application of these techniques to germplasm evaluation data, 11328 accessions of groundnut (Arachis hypogaea L) from the International Research Institute for the Semi-Arid Tropics, Andhra Pradesh, India were examined. Data for nine quantitative descriptors measured in the rainy and post-rainy growing seasons were used. The ordination technique of principal component analysis was used to reduce the dimensionality of the germplasm data. The identification of phenotypically similar groups of accessions within large scale data via the computationally intensive hierarchical clustering techniques was not feasible and non-hierarchical techniques had to be used. Finite mixture models that maximise the likelihood of an accession belonging to a cluster were used to cluster the accessions in this collection. The patterns of response for the different growing seasons were found to be highly correlated. However, in relating the results to passport and other characterisation and evaluation descriptors, the observed patterns did not appear to be related to taxonomy or any other well known characteristics of groundnut.

Evidence for consistent functional groups of wetland vegetation across a broad geographical range of Europe

We provide evidence for the existence of a general relationship between plant size and survival strategy (sensu Grime 1974) in wetland habitats across a broad geographical range in European riverine wetlands. Between April−October, in 1991 and 1992, data on eight size-related traits were collected for 144 plant populations, drawn from 85 plant species, growing in riverine wetland habitats in Spain, France, Ireland and the UK. Two non-hierarchical techniques were used to classify plant populations, on the basis of these eight traits, into three groups with a good degree of internal cohesion and external isolation.