Examining the Effectiveness of Discriminant Function Analysis and Cluster Analysis in Species Identification of Male Field Crickets Based on Their Calling Songs

Traditional taxonomy based on morphology has often failed in accurate species identification owing to the occurrence of cryptic species, which are reproductively isolated but morphologically identical. Molecular data have thus been used to complement morphology in species identification. The sexual advertisement calls in several groups of acoustically communicating animals are species-specific and can thus complement molecular data as non-invasive tools for identification. Several statistical tools and automated identifier algorithms have been used to investigate the efficiency of acoustic signals in species identification. Despite a plethora of such methods, there is a general lack of knowledge regarding the appropriate usage of these methods in specific taxa. In this study, we investigated the performance of two commonly used statistical methods, discriminant function analysis (DFA) and cluster analysis, in identification and classification based on acoustic signals of field cricket species belonging to the subfamily Gryllinae. Using a comparative approach we evaluated the optimal number of species and calling song characteristics for both the methods that lead to most accurate classification and identification. The accuracy of classification using DFA was high and was not affected by the number of taxa used. However, a constraint in using discriminant function analysis is the need for a priori classification of songs. Accuracy of classification using cluster analysis, which does not require a priori knowledge, was maximum for 6-7 taxa and decreased significantly when more than ten taxa were analysed together. We also investigated the efficacy of two novel derived acoustic features in improving the accuracy of identification. Our results show that DFA is a reliable statistical tool for species identification using acoustic signals. Our results also show that cluster analysis of acoustic signals in crickets works effectively for species classification and identification.

Incremental Linear Discriminant Analysis Using Sufficient Spanning Sets and Its Applications

This paper presents an incremental learning solution for Linear Discriminant Analysis (LDA) and its applications to object recognition problems. We apply the sufficient spanning set approximation in three steps i.e. update for the total scatter matrix, between-class scatter matrix and the projected data matrix, which leads an online solution which closely agrees with the batch solution in accuracy while significantly reducing the computational complexity. The algorithm yields an efficient solution to incremental LDA even when the number of classes as well as the set size is large. The incremental LDA method has been also shown useful for semi-supervised online learning. Label propagation is done by integrating the incremental LDA into an EM framework. The method has been demonstrated in the task of merging large datasets which were collected during MPEG standardization for face image retrieval, face authentication using the BANCA dataset, and object categorisation using the Caltech101 dataset. © 2010 Springer Science+Business Media, LLC.

On classical and other methods of discriminant analysis and estimation of log-odds

For two multinormal populations with equal covariance matrices the likelihood ratio discriminant function, an alternative allocation rule to the sample linear discriminant function when n1 ≠ n2 ,is studied analytically. With the assumption of a known covariance matrix its distribution is derived and the expectation of its actual and apparent error rates evaluated and compared with those of the sample linear discriminant function. This comparison indicates that the likelihood ratio allocation rule is robust to unequal sample sizes. The quadratic discriminant function is studied, its distribution reviewed and evaluation of its probabilities of misclassification discussed. For known covariance matrices the distribution of the sample quadratic discriminant function is derived. When the known covariance matrices are proportional exact expressions for the expectation of its actual and apparent error rates are obtained and evaluated. The effectiveness of the sample linear discriminant function for this case is also considered. Estimation of true log-odds for two multinormal populations with equal or unequal covariance matrices is studied. The estimative, Bayesian predictive and a kernel method are compared by evaluating their biases and mean square errors. Some algebraic expressions for these quantities are derived. With equal covariance matrices the predictive method is preferable. Where it derives this superiority is investigated by considering its performance for various levels of fixed true log-odds. It is also shown that the predictive method is sensitive to n1 ≠ n2. For unequal but proportional covariance matrices the unbiased estimative method is preferred. Product Normal kernel density estimates are used to give a kernel estimator of true log-odds. The effect of correlation in the variables with product kernels is considered. With equal covariance matrices the kernel and parametric estimators are compared by simulation. For moderately correlated variables and large dimension sizes the product kernel method is a good estimator of true log-odds.

A comparison of the dietary patterns derived by principal component analysis and cluster analysis in older Australians

BACKGROUND: Despite increased use of dietary pattern methods in nutritional epidemiology, there have been few direct comparisons of methods. Older adults are a particularly understudied population in the dietary pattern literature. This study aimed to compare dietary patterns derived by principal component analysis (PCA) and cluster analysis (CA) in older adults and to examine their associations with socio-demographic and health behaviours. METHODS: Men (n = 1888) and women (n = 2071) aged 55-65 years completed a 111-item food frequency questionnaire in 2010. Food items were collapsed into 52 food groups and dietary patterns were determined by PCA and CA. Associations between dietary patterns and participant characteristics were examined using Chi-square analysis. The standardised PCA-derived dietary patterns were compared across the clusters using one-way ANOVA. RESULTS: PCA identified four dietary patterns in men and two dietary patterns in women. CA identified three dietary patterns in both men and women. Men in cluster 1 (fruit, vegetables, wholegrains, fish and poultry) scored higher on PCA factor 1 (vegetable dishes, fruit, fish and poultry) and factor 4 (vegetables) compared to factor 2 (spreads, biscuits, cakes and confectionery) and factor 3 (red meat, processed meat, white-bread and hot chips) (mean, 95 % CI; 0.92, 0.82-1.02 vs. 0.74, 0.63-0.84 vs. -0.43, -0.50- -0.35 vs. 0.60 0.46-0.74, respectively). Women in cluster 1 (fruit, vegetables and fish) scored highest on PCA factor 1 (fruit, vegetables and fish) compared to factor 2 (processed meat, hot chips cakes and confectionery) (1.05, 0.97-1.14 vs. -0.14, -0.21- -0.07, respectively). Cluster 3 (small eaters) in both men and women had negative factor scores for all the identified PCA dietary patterns. Those with dietary patterns characterised by higher consumption of red and processed meat and refined grains were more likely to be Australian-born, have a lower level of education, a higher BMI, smoke and did not meet physical activity recommendations (all P < 0.05). CONCLUSIONS: PCA and CA identified comparable dietary patterns within older Australians. However, PCA may provide some advantages compared to CA with respect to interpretability of the resulting dietary patterns. Older adults with poor dietary patterns also displayed other negative lifestyle behaviours. Food-based dietary pattern methods may inform dietary advice that is understood by the community.

Identification of species of the Euterpe genus by rare earth elements using inductively coupled plasma mass spectrometry and linear discriminant analysis

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Copper ore type definition from Sossego Mine using X-ray diffraction and cluster analysis technique

This paper presents the classification of 110 copper ore samples from Sossego Mine, based on X-ray diffraction and cluster analysis. The comparison based on the position and the intensity of the diffracted peaks allowed the distinction of seven ore types, whose differences refer to the proportion of major minerals: quartz, feldspar, actinolite, iron oxides, mica and chlorite. There was a strong correlation between the grouping and the location of the samples in Sequeirinho and Sossego orebodies. This relationship is due to different types and intensities of hydrothermal alteration prevailing in each body, which reflect the mineralogical composition and thus the X-ray diffractograms of samples.

Business Clusters in Eastern Europe: Policy Analysis and Cluster Performance

Clusters have increasingly become an essential part of policy discourses at all levels, EU, national, regional, dealing with regional development, competitiveness, innovation, entrepreneurship, SMEs. These impressive efforts in promoting the concept of clusters on the policy-making arena have been accompanied by much less academic and scientific research work investigating the actual economic performance of firms in clusters, the design and execution of cluster policies and going beyond singular case studies to a more methodologically integrated and comparative approach to the study of clusters and their real-world impact. The theoretical background is far from being consolidated and there is a variety of methodologies and approaches for studying and interpreting this phenomenon while at the same time little comparability among studies on actual cluster performances. The conceptual framework of clustering suggests that they affect performance but theory makes little prediction as to the ultimate distribution of the value being created by clusters. This thesis takes the case of Eastern European countries for two reasons. One is that clusters, as coopetitive environments, are a new phenomenon as the previous centrally-based system did not allow for such types of firm organizations. The other is that, as new EU member states, they have been subject to the increased popularization of the cluster policy approach by the European Commission, especially in the framework of the National Reform Programmes related to the Lisbon objectives. The originality of the work lays in the fact that starting from an overview of theoretical contributions on clustering, it offers a comparative empirical study of clusters in transition countries. There have been very few examples in the literature that attempt to examine cluster performance in a comparative cross-country perspective. It adds to this an analysis of cluster policies and their implementation or lack of such as a way to analyse the way the cluster concept has been introduced to transition economies. Our findings show that the implementation of cluster policies does vary across countries with some countries which have embraced it more than others. The specific modes of implementation, however, are very similar, based mostly on soft measures such as funding for cluster initiatives, usually directed towards the creation of cluster management structures or cluster facilitators. They are essentially founded on a common assumption that the added values of clusters is in the creation of linkages among firms, human capital, skills and knowledge at the local level, most often perceived as the regional level. Often times geographical proximity is not a necessary element in the application process and cluster application are very similar to network membership. Cluster mapping is rarely a factor in the selection of cluster initiatives for funding and the relative question about critical mass and expected outcomes is not considered. In fact, monitoring and evaluation are not elements of the cluster policy cycle which have received a lot of attention. Bulgaria and the Czech Republic are the countries which have implemented cluster policies most decisively, Hungary and Poland have made significant efforts, while Slovakia and Romania have only sporadically and not systematically used cluster initiatives. When examining whether, in fact, firms located within regional clusters perform better and are more efficient than similar firms outside clusters, we do find positive results across countries and across sectors. The only country with negative impact from being located in a cluster is the Czech Republic.