Random projections on manifolds of symmetric positive definite matrices for image classification

Recent advances suggest that encoding images through Symmetric Positive Definite (SPD) matrices and then interpreting such matrices as points on Riemannian manifolds can lead to increased classification performance. Taking into account manifold geometry is typically done via (1) embedding the manifolds in tangent spaces, or (2) embedding into Reproducing Kernel Hilbert Spaces (RKHS). While embedding into tangent spaces allows the use of existing Euclidean-based learning algorithms, manifold shape is only approximated which can cause loss of discriminatory information. The RKHS approach retains more of the manifold structure, but may require non-trivial effort to kernelise Euclidean-based learning algorithms. In contrast to the above approaches, in this paper we offer a novel solution that allows SPD matrices to be used with unmodified Euclidean-based learning algorithms, with the true manifold shape well-preserved. Specifically, we propose to project SPD matrices using a set of random projection hyperplanes over RKHS into a random projection space, which leads to representing each matrix as a vector of projection coefficients. Experiments on face recognition, person re-identification and texture classification show that the proposed approach outperforms several recent methods, such as Tensor Sparse Coding, Histogram Plus Epitome, Riemannian Locality Preserving Projection and Relational Divergence Classification.

BIM, specifications and cost planning: establishing the connections

The overarching research work is based on two approaches: - Conceptual Analysis, Extraction and Linking - Experimentation with Product Libraries - Conceptual Analysis, Extraction and Linking: This aspect of the research has been achieved through the development of a conceptual framework for facilitating the understanding of the constituting components of BIM, Specifications and Cost Planning under investigation. The framework builds on theories spanning the constituent research themes and was used as a basis for justifying the elected approaches adopted throughout the research work. By means of tags and codes, a system for classifying building specification information has been developed as a differentiator between the chosen research approach and existing classification strategies in industry. Furthermore, syntactic links between extracted classes of specification information and cost planning have been established and will be adopted as a basis for authenticating the impact of specification information within BIM models. - Experimentation with Product Libraries Following the extraction and classification of BIM, Specifications and Cost Planning information, early experimentation on linking specifications to BIM models by means of a raas-based product library have been successful. A comparative analysis between a range of existing product libraries has also been realised. The outcomes have been amply documented in papers, all of which have received positive reviews. Ongoing experiments and analysis with the product library involve integrating the cost planning component for authenticating the completeness, relevance and impact of embedded specification within BIM models.

Signaling pathway genes for blood pressure, folate and cholesterol levels among hypertensives : an epistasis analysis

Irregular atrial pressure, defective folate and cholesterol metabolism contribute to the pathogenesis of hypertension. However, little is known about the combined roles of the methylenetetrahydrofolate reductase (MTHFR), apolipoprotein-E (ApoE) and angiotensin-converting enzyme (ACE) genes, which are involved in metabolism and homeostasis. The objective of this study is to investigate the association of the MTHFR 677 C>T and 1298A>C, ACE insertion–deletion (I/D) and ApoE genetic polymorphisms with hypertension and to further explore the epistasis interactions that are involved in these mechanisms. A total of 594 subjects, including 348 normotensive and 246 hypertensive ischemic stroke subjects were recruited. The MTHFR 677 C>T and 1298A>C, ACE I/D and ApoEpolymorphisms were genotyped and the epistasis interaction were analyzed. The MTHFR 677 C>T and ApoE polymorphisms demonstrated significant associations with susceptibility to hypertension in multiple logistic regression models, multifactor dimensionality reduction and a classification and regression tree. In addition, the logistic regression model demonstrated that significant interactions between the ApoE E3E3, E2E4, E2E2 and MTHFR 677 C>T polymorphisms existed. In conclusion, the results of this epistasis study indicated significant association between the ApoE and MTHFR polymorphisms and hypertension.

“I would have lost the respect of my friends and family if they knew I had bent the road rules” : parents, peers, and the perilous behaviour of young drivers

Young novice drivers are at considerable risk of injury on the road. Their behaviour appears vulnerable to the social influence of their parents and friends. The nature and mechanisms of parent and peer influence on young novice driver (16–25 years) behaviour was explored via small group interviews (n = 21) and two surveys (n1 = 1170, n2 = 390) to inform more effective young driver countermeasures. Parental and peer influence occurred in preLicence, Learner, and Provisional (intermediate) periods. Pre-Licence and unsupervised Learner drivers reported their parents were less likely to punish risky driving (e.g., speeding). These drivers were more likely to imitate their parents and reported their parents were also risky drivers. Young novice drivers who experienced or expected social punishments from peers, including ‘being told off’ for risky driving, reported less riskiness. Conversely drivers who experienced or expected social rewards such as being ‘cheered on’ by friends – who were also more risky drivers – reported more risky driving including crashes and offences. Interventions enhancing positive influence and curtailing negative influence may improve road safety outcomes not only for young novice drivers, but for all persons who share the road with them. Parent-specific interventions warrant further development and evaluation including: modelling safe driving behaviour by parents; active monitoring of driving during novice licensure; and sharing the family vehicle during the intermediate phase. Peer-targeted interventions including modelling of safe driving behaviour and attitudes; minimisation of social reinforcement and promotion of social sanctions for risky driving also need further development and evaluation.

Human vs. machine : identification of bat species from their echolocation calls by humans and by artificial neural networks

Automated remote ultrasound detectors allow large amounts of data on bat presence and activity to be collected. Processing of such data involves identifying bat species from their echolocation calls. Automated species identification has the potential to provide more consistent, predictable, and potentially higher levels of accuracy than identification by humans. In contrast, identification by humans permits flexibility and intelligence in identification, as well as the incorporation of features and patterns that may be difficult to quantify. We compared humans with artificial neural networks (ANNs) in their ability to classify short recordings of bat echolocation calls of variable signal to noise ratios; these sequences are typical of those obtained from remote automated recording systems that are often used in large-scale ecological studies. We presented 45 recordings (1–4 calls) produced by known species of bats to ANNs and to 26 human participants with 1 month to 23 years of experience in acoustic identification of bats. Humans correctly classified 86% of recordings to genus and 56% to species; ANNs correctly identified 92% and 62%, respectively. There was no significant difference between the performance of ANNs and that of humans, but ANNs performed better than about 75% of humans. There was little relationship between the experience of the human participants and their classification rate. However, humans with <1 year of experience performed worse than others. Currently, identification of bat echolocation calls by humans is suitable for ecological research, after careful consideration of biases. However, improvements to ANNs and the data that they are trained on may in future increase their performance to beyond those demonstrated by humans.

A Boundary-Oriented Chinese Segmentation Method Using NGram Mutual Information

This paper describes our participation in the Chinese word segmentation task of CIPS-SIGHAN 2010. We implemented an n-gram mutual information (NGMI) based segmentation algorithm with the mixed-up features from unsupervised, supervised and dictionarybased segmentation methods. This algorithm is also combined with a simple strategy for out-of-vocabulary (OOV) word recognition. The evaluation for both open and closed training shows encouraging results of our system. The results for OOV word recognition in closed training evaluation were however found unsatisfactory.

Classification of inventory data suit for managing Australian railway bridges

With a focus to optimising the life cycle performance of Australian Railway bridges, new bridge classification and environmental classification systems are proposed. The new bridge classification system is mainly to facilitate the implementation of novel Bridge Management System (BMS) which optimise the life cycle cost both at project level and network level while environment classification is mainly to improve accuracy of Remaining Service Potential (RSP) module of the proposed BMS. In fact, limited capacity of the existing BMS to trigger the maintenance intervention point is an indirect result of inadequacies of the existing bridge and environmental classification systems. The proposed bridge classification system permits to identify the intervention points based on percentage deterioration of individual elements and maintenance cost, while allowing performance based rating technique to implement for maintenance optimisation and prioritisation. Simultaneously, the proposed environment classification system will enhance the accuracy of prediction of deterioration of steel components.

Assistive classification for improving the efficiency of avian species richness surveys

Avian species richness surveys, which measure the total number of unique avian species, can be conducted via remote acoustic sensors. An immense quantity of data can be collected, which, although rich in useful information, places a great workload on the scientists who manually inspect the audio. To deal with this big data problem, we calculated acoustic indices from audio data at a one-minute resolution and used them to classify one-minute recordings into five classes. By filtering out the non-avian minutes, we can reduce the amount of data by about 50% and improve the efficiency of determining avian species richness. The experimental results show that, given 60 one-minute samples, our approach enables to direct ecologists to find about 10% more avian species.

Taxonomic revision of Ustilago, Sporisorium and Macalpinomyces

Morphological characters within the Ustilago-Sporisorium-Macalpinomyces complex are defined explicitly. The genera Sporisorium and Anthracocystis are emended to reflect morphological synapomorphies. Three new genera, Langdonia, Stollia and Triodiomyces are described based on soral synapomorphies and host classification. The new classification of the Ustilago-Sporisorium-Macalpinomyces complex incorporates 142 new taxonomic combinations.

Host Rock Classification (HRC) system for nuclear waste disposal in crystalline bedrock

A new rock mass classification scheme, the Host Rock Classification system (HRC-system) has been developed for evaluating the suitability of volumes of rock mass for the disposal of high-level nuclear waste in Precambrian crystalline bedrock. To support the development of the system, the requirements of host rock to be used for disposal have been studied in detail and the significance of the various rock mass properties have been examined. The HRC-system considers both the long-term safety of the repository and the constructability in the rock mass. The system is specific to the KBS-3V disposal concept and can be used only at sites that have been evaluated to be suitable at the site scale. By using the HRC-system, it is possible to identify potentially suitable volumes within the site at several different scales (repository, tunnel and canister scales). The selection of the classification parameters to be included in the HRC-system is based on an extensive study on the rock mass properties and their various influences on the long-term safety, the constructability and the layout and location of the repository. The parameters proposed for the classification at the repository scale include fracture zones, strength/stress ratio, hydraulic conductivity and the Groundwater Chemistry Index. The parameters proposed for the classification at the tunnel scale include hydraulic conductivity, Q´ and fracture zones and the parameters proposed for the classification at the canister scale include hydraulic conductivity, Q´, fracture zones, fracture width (aperture + filling) and fracture trace length. The parameter values will be used to determine the suitability classes for the volumes of rock to be classified. The HRC-system includes four suitability classes at the repository and tunnel scales and three suitability classes at the canister scale and the classification process is linked to several important decisions regarding the location and acceptability of many components of the repository at all three scales. The HRC-system is, thereby, one possible design tool that aids in locating the different repository components into volumes of host rock that are more suitable than others and that are considered to fulfil the fundamental requirements set for the repository host rock. The generic HRC-system, which is the main result of this work, is also adjusted to the site-specific properties of the Olkiluoto site in Finland and the classification procedure is demonstrated by a test classification using data from Olkiluoto. Keywords: host rock, classification, HRC-system, nuclear waste disposal, long-term safety, constructability, KBS-3V, crystalline bedrock, Olkiluoto

Complete genome sequence and intracellular protein localization of Datura yellow vein nucleorhabdovirus

A limited number of plant rhabdovirus genomes have been fully sequenced, making taxonomic classification, evolutionary analysis and molecular characterization of this virus group difficult. We have for the first time determined the complete genome sequence of 13,188 nucleotides of Datura yellow vein nucleorhabdovirus (DYVV). DYVV genome organization resembles that of its closest relative, Sonchus yellow net virus (SYNV), with six ORFs in antigenomic orientation, separated by highly conserved intergenic regions and flanked by complementary 3′ leader and 5′ trailer sequences. As is typical for nucleorhabdoviruses, all viral proteins, except the glycoprotein, which is targeted to the endoplasmic reticulum, are localized to the nucleus. Nucleocapsid (N) protein, matrix (M) protein and polymerase, as components of nuclear viroplasms during replication, have predicted strong canonical nuclear localization signals, and N and M proteins exclusively localize to the nucleus when transiently expressed as GFP fusions. As in all nucleorhabdoviruses studied so far, N and phosphoprotein P interact when co-expressed, significantly increasing P nuclear localization in the presence of N protein. This research adds to the list of complete genomes of plant-infecting rhabdoviruses, provides molecular tools for further characterization and supports classification of DYVV as a nucleorhabdovirus closely related to but with some distinct differences from SYNV.

Männyn neulaskadon arviointi laserkeilauksella

Leaf and needle biomasses are key factors in forest health. Insects that feed on needles cause growth losses and tree mortality. Insect outbreaks in Finnish forests have increased rapidly during the last decade and due to climate change the damages are expected to become more serious. There is a need for cost-efficient methods for inventorying these outbreaks. Remote sensing is a promising means for estimating forests and damages. The purpose of this study is to investigate the usability of airborne laser scanning in estimating Scots pine defoliation caused by the common pine sawfly (Diprion pini L.). The study area is situated in Ilomantsi district, eastern Finland. Study materials included high-pulse airborne laser scannings from July and October 2008. Reference data consisted of 90 circular field plots measured in May-June 2009. Defoliation percentage on these field plots was estimated visually. The study was made on plot-level and methods used were linear regression, unsupervised classification, Maximum likelihood method, and stepwise linear regression. Field plots were divided in defoliation classes in two different ways: When divided in two classes the defoliation percentages used were 0–20 % and 20–100 % and when divided in four classes 0–10 %, 10–20 %, 20–30 % and 30–100 %. The results varied depending on method and laser scanning. In the first laser scanning the best results were obtained with stepwise linear regression. The kappa value was 0,47 when using two classes and 0,37 when divided in four classes. In the second laser scanning the best results were obtained with Maximum likelihood. The kappa values were 0,42 and 0,37, correspondingly. The feature that explained defoliation best was vegetation index (pulses reflected from height > 2m / all pulses). There was no significant difference in the results between the two laser scannings so the seasonal change in defoliation could not be detected in this study.

A Framework for Classification of Prokaryotic Protein Kinases

Background:Overwhelming majority of the Serine/Threonine protein kinases identified by gleaning archaeal and eubacterial genomes could not be classified into any of the well known Hanks and Hunter subfamilies of protein kinases. This is owing to the development of Hanks and Hunter classification scheme based on eukaryotic protein kinases which are highly divergent from their prokaryotic homologues. A large dataset of prokaryotic Serine/Threonine protein kinases recognized from genomes of prokaryotes have been used to develop a classification framework for prokaryotic Ser/Thr protein kinases. Methodology/Principal Findings: We have used traditional sequence alignment and phylogenetic approaches and clustered the prokaryotic kinases which represent 72 subfamilies with at least 4 members in each. Such a clustering enables classification of prokaryotic Ser/Thr kinases and it can be used as a framework to classify newly identified prokaryotic Ser/Thr kinases. After series of searches in a comprehensive sequence database we recognized that 38 subfamilies of prokaryotic protein kinases are associated to a specific taxonomic level. For example 4, 6 and 3 subfamilies have been identified that are currently specific to phylum proteobacteria, cyanobacteria and actinobacteria respectively. Similarly subfamilies which are specific to an order, sub-order, class, family and genus have also been identified. In addition to these, we also identify organism-diverse subfamilies. Members of these clusters are from organisms of different taxonomic levels, such as archaea, bacteria, eukaryotes and viruses.Conclusion/Significance: Interestingly, occurrence of several taxonomic level specific subfamilies of prokaryotic kinases contrasts with classification of eukaryotic protein kinases in which most of the popular subfamilies of eukaryotic protein kinases occur diversely in several eukaryotes. Many prokaryotic Ser/Thr kinases exhibit a wide variety of modular organization which indicates a degree of complexity and protein-protein interactions in the signaling pathways in these microbes.

Decision tree design and applications in speech processing

The design and operation of the minimum cost classifier, where the total cost is the sum of the measurement cost and the classification cost, is computationally complex. Noting the difficulties associated with this approach, decision tree design directly from a set of labelled samples is proposed in this paper. The feature space is first partitioned to transform the problem to one of discrete features. The resulting problem is solved by a dynamic programming algorithm over an explicitly ordered state space of all outcomes of all feature subsets. The solution procedure is very general and is applicable to any minimum cost pattern classification problem in which each feature has a finite number of outcomes. These techniques are applied to (i) voiced, unvoiced, and silence classification of speech, and (ii) spoken vowel recognition. The resulting decision trees are operationally very efficient and yield attractive classification accuracies.

Gene expression programming-fuzzy logic method for crop type classification

Crop type classification using remote sensing data plays a vital role in planning cultivation activities and for optimal usage of the available fertile land. Thus a reliable and precise classification of agricultural crops can help improve agricultural productivity. Hence in this paper a gene expression programming based fuzzy logic approach for multiclass crop classification using Multispectral satellite image is proposed. The purpose of this work is to utilize the optimization capabilities of GEP for tuning the fuzzy membership functions. The capabilities of GEP as a classifier is also studied. The proposed method is compared to Bayesian and Maximum likelihood classifier in terms of performance evaluation. From the results we can conclude that the proposed method is effective for classification.