Temporal extension of Laplacian Eigenmaps for unsupervised dimensionality reduction of time series

A novel non-linear dimensionality reduction method, called Temporal Laplacian Eigenmaps, is introduced to process efficiently time series data. In this embedded-based approach, temporal information is intrinsic to the objective function, which produces description of low dimensional spaces with time coherence between data points. Since the proposed scheme also includes bidirectional mapping between data and embedded spaces and automatic tuning of key parameters, it offers the same benefits as mapping-based approaches. Experiments on a couple of computer vision applications demonstrate the superiority of the new approach to other dimensionality reduction method in term of accuracy. Moreover, its lower computational cost and generalisation abilities suggest it is scalable to larger datasets. © 2010 IEEE.

Online Multiperson Tracking With Occlusion Reasoning and Unsupervised Track Motion Model

We address the problem of multi-target tracking in realistic crowded conditions by introducing a novel dual-stage online tracking algorithm. The problem of data-association between tracks and detections, based on appearance, is often complicated by partial occlusion. In the first stage, we address the issue of occlusion with a novel method of robust data-association, that can be used to compute the appearance similarity between tracks and detections without the need for explicit knowledge of the occluded regions. In the second stage, broken tracks are linked based on motion and appearance, using an online-learned linking model. The online-learned motion-model for track linking uses the confident tracks from the first stage tracker as training examples. The new approach has been tested on the town centre dataset and has performance comparable with the present state-of-the-art

Unsupervised Solution Post Identification from Discussion Forums

Discussion forums have evolved into a dependablesource of knowledge to solvecommon problems. However, only a minorityof the posts in discussion forumsare solution posts. Identifying solutionposts from discussion forums, hence, is animportant research problem. In this paper,we present a technique for unsupervisedsolution post identification leveraginga so far unexplored textual feature, thatof lexical correlations between problemsand solutions. We use translation modelsand language models to exploit lexicalcorrelations and solution post characterrespectively. Our technique is designedto not rely much on structural featuressuch as post metadata since suchfeatures are often not uniformly availableacross forums. Our clustering-based iterativesolution identification approach basedon the EM-formulation performs favorablyin an empirical evaluation, beatingthe only unsupervised solution identificationtechnique from literature by a verylarge margin. We also show that our unsupervisedtechnique is competitive againstmethods that require supervision, outperformingone such technique comfortably.

Randomised controlled trial of home-based walking programmes at and below current recommended levels of exercise in sedentary adults

Objectives: To determine, using unsupervised walking programmes, the effects of exercise at a level lower than currently recommended to improve cardiovascular risk factors and functional capacity.

MISEP method for postnonlinear blind source separation

In this letter, a standard postnonlinear blind source separation algorithm is proposed, based on the MISEP method, which is widely used in linear and nonlinear independent component analysis. To best suit a wide class of postnonlinear mixtures, we adapt the MISEP method to incorporate a priori information of the mixtures. In particular, a group of three-layered perceptrons and a linear network are used as the unmixing system to separate sources in the postnonlinear mixtures, and another group of three-layered perceptron is used as the auxiliary network. The learning algorithm for the unmixing system is then obtained by maximizing the output entropy of the auxiliary network. The proposed method is applied to postnonlinear blind source separation of both simulation signals and real speech signals, and the experimental results demonstrate its effectiveness and efficiency in comparison with existing methods.

Home curation versus teenage photography: Photo displays in the family home

In this paper we report an empirical study of the photographic portrayal of family members at home. Adopting a social psychological approach and focusing oil intergenerational power dynamics, our research explores the use of domestic photo displays in family representation. Parents and their teenagers from eight families in the south of England were interviewed at home about their interpretations of both stored and displayed photos within the home. Discussions centred on particular photographs found by the participants to portray self and family in different ways. The findings show that public displays of digital photos are still curated by mothers of the households, but with more difficulty and less control all with analogue photos. In addition, teenagers both contribute and comply with this curation within the home, whilst at the same time developing additional ways of presenting their families and themselves online that are 'unsupervised' by the curator. We highlight the conflict of interest that is at play within teen and parent practices and consider the challenges that this presents for supporting the representation of family through the design of photo display technology. (C) 2009 Elsevier Ltd. All rights reserved.

Impact of Sample Heterogeneity on Methylation Analysis

The recent emergence of high-throughput arrays for methylation analysis has made the influence of tumor content on the interpretation of methylation levels increasingly pertinent. However, to what degree does tumor content have an influence, and what degree of tumor content makes a specimen acceptable for accurate analysis remains unclear. Taking a systematic approach, we analyzed 98 unselected formalin-fixed and paraffin-embedded gastric tumors and matched normal tissue samples using the Illumina GoldenGate methylation assay. Unsupervised hierarchical clustering showed 2 separate clusters with a significant difference in average tumor content levels. The probes identified to be significantly differentially methylated between the tumors and normals also differed according to the tumor content of the samples included, with the sensitivity of identifying the

Tissue microarrays characterise the clinical significance of a VEGF-A protein expression signature in gastrointestinal stromal tumours

A tissue microarray analysis of 22 proteins in gastrointestinal stromal tumours ( GIST), followed by an unsupervised, hierarchical monothetic cluster statistical analysis of the results, allowed us to detect a vascular endothelial growth factor ( VEGF) protein overexpression signature discriminator of prognosis in GIST, and discover novel VEGF-A DNA variants that may have functional significance.

An international standardization programme towards the application of gene expression profiling in routine leukaemia diagnostics: the Microarray Innovations in LEukemia study prephase

Gene expression profiling has the potential to enhance current methods for the diagnosis of haematological malignancies. Here, we present data on 204 analyses from an international standardization programme that was conducted in 11 laboratories as a prephase to the Microarray Innovations in LEukemia (MILE) study. Each laboratory prepared two cell line samples, together with three replicate leukaemia patient lysates in two distinct stages: (i) a 5-d course of protocol training, and (ii) independent proficiency testing. Unsupervised, supervised, and r(2) correlation analyses demonstrated that microarray analysis can be performed with remarkably high intra-laboratory reproducibility and with comparable quality and reliability.

Exercise and Auricular Acupuncture for Chronic Low Back Pain: A Feasibility Randomised Controlled Trial

Objectives: To evaluate the feasibility of a randomized-controlled trial (RCT) investigating the effects of adding auricular acupuncture (AA) to exercise for participants with chronic low-back pain (CLBP). Methods: Participants with CLBP were recruited from primary care and a university population and were randomly allocated (n=51) to 1 of 2 groups: (1) "Exercise Alone (E)"-12-week program consisting of 6 weeks of supervised exercise followed by 6 weeks unsupervised exercise (n=27); or (2) "Exercise and AA (EAA)"-12-week exercise program and AA (n=24). Outcome measures were recorded at baseline, week 8, week 13, and 6 months. The primary outcome measure was the Oswestry Disability Questionnaire. Results: Participants in the EAA group demonstrated a greater mean improvement of 10.7% points (95% confidence interval, -15.3,-5.7) (effect size=1.20) in the Oswestry Disability Questionnaire at 6 months compared with 6.7% points (95% confidence interval, -11.4,-1.9) in the E group (effect size=0.58). There was also a trend towards a greater mean improvement in quality of life, LBP intensity and bothersomeness, and fear-avoidance beliefs in the EAA group. The dropout rate for this trial was lower than anticipated (15% at 6 mo), adherence with exercise was similar (72% E; 65% EAA). Adverse effects for AA ranged from 1% to 14% of participants. Discussion: Findings of this study showed that a main RCT is feasible and that 56 participants per group would need to be recruited, using multiple recruitment approaches. AA was safe and demonstrated additional benefits when combined with exercise for people with CLBP, which requires confirmation in a fully powered RCT.