In this study, a new artificial intelligence optimization algorithm, Differential Search (DS), was proposed for Principal Component Analysis (PCA) based unsupervised change detection method for optic and SAR image data. The model firstly computes an eigenvector space using previously created k×k blocks. The change detection map is generated by clustering the feature vector as two clusters which are changed and unchanged using Differential Search Algorithm. For clustering, a cost function is used based on minimization of Euclidean distance between cluster centers and pixels. Experimental results of optic and SAR images proved that proposed approach is effective for unsupervised change detection of remote sensing image data.
Linear discriminant analysis (LDA) is a versatile method in all pattern recognition fields but it suffers from some limitations. In a multi-class problem, when samples of a class are far from other classes samples, it leads to bias of the whole decision boundaries of LDA in favor of the farthest class. To overcome this drawback, this study is aimed at minimizing this bias by redefining the between- and within-class scatter matrices via incorporating weight vectors derived from Fisher value of classes pairs. After projecting the input patterns into a lower-dimensional space in which the class samples are more separable, a new version of nearest neighbor (NN) method with an adaptive distance measure is employed to classify the transformed samples. To speed up the adaptive distance routine, an iterative learning algorithm that minimizes the error rate is presented. This efficient method is applied to six standard datasets driven from the UCI repository dataset and test results are evaluated from three aspects in terms of accuracy, robustness, and complexity. Results show the supremacy of the proposed two-layer classifier in comparison with the combination of different versions of LDA and NN methods from the three points of view. Moreover, the proposed classifier is assessed in the noisy environment of those datasets and the achieved results confirm the high robustness of the introduced scheme when compared to others.
We assessed the xtent of temporal variation and autocorrelation in fish habitat use based on an experimental study of individual 0+ juvenile barbel, Barbus barbus, in an artificial flume. Five treated and five control fish were individually subjected to an increase in discharege (intervention) hlfway through each experiment and kept at baseline discharge throughout, respectively. Preference surves for velocity were generated for each of 60 trials per experiment and for each combination of treated/control (fish) x before/after-intervention. There were large between- and within-individual differences in velocity preference, both in treated and in control fish. Most barbel explored the entire range of velocities, whereas some individuals used a more limited range. Temporal variation in behavioural responses was assessed by a PCA-based methodology. Autocorrelation (i.e. correlation between sequential trials) was diagnosed in most response profiles, supporting recent fidings that individuals may have a "memory" of their past velocity usage. The relevance of the results for numerical habitat models of fish habitat assessment is discussed, as well as the importance of incorporating temporal variability into fish habitat use models (e.g. PHABSIM), not only as ontogenetic intervals but also as longitudinal data of individual behaviours. A warning is also re-issued about the erroneous belief of "pseudoreplication" simly arising from repeated measurements in time.
Community structure implies some features in various real-world networks, and these features can help us to analysis structural and functional properties in the complex system. It has been proved that the classic k-means algorithm can efficiently cluster nodes into communities. However, initial seeds decide the efficiency of the k-means, especially when detecting communities with different sizes. To solve this problem, we improve the classic community detection algorithm with Principal Component Analysis (PCA) mapping and local expansion k-means. Since PCA can preserve the distance information of every node pairs, the improved algorithm use PCA to map nodes in the complex network into lower dimension European space, and then detect initial seeds for k-means using the improved local expansion strategy. Based on the chosen initial seeds, the k-means algorithm can cluster nodes into communities. We apply the proposed algorithm in real-world and artificial networks, the results imply that the improved algorithm is efficient to detect communities and is robust to the initial seed of K-means.
Slow fluctuating radar targets have shown to be very difficult to classify by means of neural networks. This paper deals with the application of time-frequency decompositions for improving the performance of neural networks for this kind of targets. Several topics, such as dimensionality reduction of the time-frequency representations and the optimum value of SNR for training are discussed. The proposed detector is compared with a single neural network for radar detection, showing that he performance is improved for slow fluctuating radar targets, especially for low values of the probability of false alarm.
Příspěvek pojednává o psychoterapeutickém směru PCA (Person-Centered Approach) - „přístup zaměřený na člověka", který vytvořil C. R. Rogers. Autorka si pokládá otázky po účinnosti tohoto přístupu a odpovídá na ně dvěma kazuistikami vybranými ze své psychoterapeutické praxe, v níž se věnuje převážně klientům s psychiatrickou diagnózou. and The paper is concerned with the PCA (Person-Centered Approach), a psychotherapeutic approach developed by C.R. Rogers. Examinig efficiency of this approach, the author is answering by two case reports from her own psychotherapeutical praxis which is aimed mainly at clients with psychiatric diagnosis.
Quantifying the functional diversity in ecological communities is very promising for both studying the response of diversity to environmental gradients and the effects of diversity on ecosystem functioning (i.e. in “biodiversity experiments”). In our view, the Rao coefficient is a good candidate for an efficient functional diversity index. It is, in fact, a generalization of the Simpson’s index of diversity and it can be used with various measures of dissimilarity between species (both those based on a single trait and those based on several traits). However, when intending to quantify the functional diversity, we have to make various methodological decisions such as how many and which traits to use, how to weight them, how to combine traits that are measured at different scales and how to quantify the species’ relative abundances in a community. Here we discuss these issues with examples from real plant communities and argue that diversity within a single trait is often the most ecologically relevant information. When using indices based on many traits, we plead for careful a priori selection of ecologically relevant traits, although other options are also feasible. When combining many traits, often with different scales, methods considering the extent of species overlap in trait space can be applied for both the qualitative and quantitative traits. Another possibility proposed here is to decompose the variability of a trait in a community according to the relative effect of among- and within-species differentiation (with the latter not considered by current indices of functional diversity), in a way analogical to decomposition of Sum of squares in ANOVA. Further, we show why the functional diversity is more tightly related to species diversity (measured by Simpson index) when biomass is used as a measure of population abundance, in comparison with frequency. Finally, the general expectation is that functional diversity can be a better predictor of ecosystem functioning than the number of species or the number of functional groups. However, we demonstrate that some of the expectations might be overrated – in particular, the “sampling effect“ in biodiversity experiments is not avoided when functional diversity is used as a predictor.
The species richness of free-living vertebrates was analysed using mapping of occurrence within individual grid squares (12 x 11.1 km) over the territory of the Czech Republic. The data on species distribution were derived from recent distributional atlases published in the last 15 years, and the records originated mostly in the last 20 years. Altogether, 384 species of cyclostomes, bony fishes, amphibians, reptiles, birds and mammals were included in this study and their presence or absence was recorded in 678 grid squares. The species numbers ascertained in the 523 grid squares situated completely within the Czech Republic varied from 92 to 259 species, with a median of 182 species. The first two principal components explained 44.9 % of the total variance and separated two main habitat gradients based on values of different environmental, topographic, and demographic variables in particular squares. The PC1 represents a gradient from urban habitats at lower altitudes to more homogenous habitats with dominant coniferous forests and meadows situated at higher altitudes. The importance of natural habitats (represented by broad-leaved and mixed forests, as well as by protected areas) and landscape heterogeneity increases along the PC2. Generalized Linear Modelling for each group of vertebrates was fitted using the number of species of individual vertebrate groups as a response variable and the first two principal components as explanatory variables. The species richness of all vertebrate groups except for reptiles is highly dependent on the PC1. The number of fish, amphibian, and bird species in squares decreases with increasing value of the PC1, i.e. it is higher in urban areas at lower altitudes. By contrast, the number of mammal species is higher in uninhabited areas at higher altitudes. The gradient represented by the PC2 is highly significant for species richness of reptiles and mammals, and the number of species of both groups increases with increasing importance of natural habitats.