Deep learning via semi-supervised embedding

We show how nonlinear embedding algorithms popular for use with shallow semi-supervised learning techniques such as kernel methods can be applied to deep multilayer architectures, either as a regularizer at the output layer, or on each layer of the architecture. This provides a simple alternative to existing approaches to deep learning whilst yielding competitive error rates compared to those methods, and existing shallow semi-supervised techniques.

A Survey of Active Learning Algorithms for Supervised Remote Sensing Image Classification

Defining an efficient training set is one of the most delicate phases for the success of remote sensing image classification routines. The complexity of the problem, the limited temporal and financial resources, as well as the high intraclass variance can make an algorithm fail if it is trained with a suboptimal dataset. Active learning aims at building efficient training sets by iteratively improving the model performance through sampling. A user-defined heuristic ranks the unlabeled pixels according to a function of the uncertainty of their class membership and then the user is asked to provide labels for the most uncertain pixels. This paper reviews and tests the main families of active learning algorithms: committee, large margin, and posterior probability-based. For each of them, the most recent advances in the remote sensing community are discussed and some heuristics are detailed and tested. Several challenging remote sensing scenarios are considered, including very high spatial resolution and hyperspectral image classification. Finally, guidelines for choosing the good architecture are provided for new and/or unexperienced user.

Semi-supervised segmentation of ultrasound images based on patch representation and continuous min cut.

Ultrasound segmentation is a challenging problem due to the inherent speckle and some artifacts like shadows, attenuation and signal dropout. Existing methods need to include strong priors like shape priors or analytical intensity models to succeed in the segmentation. However, such priors tend to limit these methods to a specific target or imaging settings, and they are not always applicable to pathological cases. This work introduces a semi-supervised segmentation framework for ultrasound imaging that alleviates the limitation of fully automatic segmentation, that is, it is applicable to any kind of target and imaging settings. Our methodology uses a graph of image patches to represent the ultrasound image and user-assisted initialization with labels, which acts as soft priors. The segmentation problem is formulated as a continuous minimum cut problem and solved with an efficient optimization algorithm. We validate our segmentation framework on clinical ultrasound imaging (prostate, fetus, and tumors of the liver and eye). We obtain high similarity agreement with the ground truth provided by medical expert delineations in all applications (94% DICE values in average) and the proposed algorithm performs favorably with the literature.

Biological embedding of early life exposures and disease risk in humans: a role for DNA methylation.

BACKGROUND: Following wider acceptance of "the thrifty phenotype" hypothesis and the convincing evidence that early life exposures can influence adult health even decades after the exposure, much interest has been placed on the mechanisms through which early life exposures become biologically embedded. METHODS: In this review, we summarize the current literature regarding biological embedding of early life experiences. To this end we conducted a literature search to identify studies investigating early life exposures in relation to DNA methylation changes. In addition, we summarize the challenges faced in investigations of epigenetic effects, stemming from the peculiarities of this emergent and complex field. A proper systematic review and meta-analyses were not feasible given the nature of the evidence. RESULTS: We identified 7 studies on early life socioeconomic circumstances, 10 studies on childhood obesity, and 6 studies on early life nutrition all relating to DNA methylation changes that met the stipulated inclusion criteria. The pool of evidence gathered, albeit small, favours a role of epigenetics and DNA methylation in biological embedding, but replication of findings, multiple comparison corrections, publication bias, and causality are concerns remaining to be addressed in future investigations. CONCLUSIONS: Based on these results, we hypothesize that epigenetics, in particular DNA methylation, is a plausible mechanism through which early life exposures are biologically embedded. This review describes the current status of the field and acts as a stepping stone for future, better designed investigations on how early life exposures might become biologically embedded through epigenetic effects. This article is protected by copyright. All rights reserved.

Uncertainty Quantification Of A Semi-Supervised Support Vector Regression Reservoir Model

Uncertainty quantification of petroleum reservoir models is one of the present challenges, which is usually approached with a wide range of geostatistical tools linked with statistical optimisation or/and inference algorithms. Recent advances in machine learning offer a novel approach to model spatial distribution of petrophysical properties in complex reservoirs alternative to geostatistics. The approach is based of semisupervised learning, which handles both ?labelled? observed data and ?unlabelled? data, which have no measured value but describe prior knowledge and other relevant data in forms of manifolds in the input space where the modelled property is continuous. Proposed semi-supervised Support Vector Regression (SVR) model has demonstrated its capability to represent realistic geological features and describe stochastic variability and non-uniqueness of spatial properties. On the other hand, it is able to capture and preserve key spatial dependencies such as connectivity of high permeability geo-bodies, which is often difficult in contemporary petroleum reservoir studies. Semi-supervised SVR as a data driven algorithm is designed to integrate various kind of conditioning information and learn dependences from it. The semi-supervised SVR model is able to balance signal/noise levels and control the prior belief in available data. In this work, stochastic semi-supervised SVR geomodel is integrated into Bayesian framework to quantify uncertainty of reservoir production with multiple models fitted to past dynamic observations (production history). Multiple history matched models are obtained using stochastic sampling and/or MCMC-based inference algorithms, which evaluate posterior probability distribution. Uncertainty of the model is described by posterior probability of the model parameters that represent key geological properties: spatial correlation size, continuity strength, smoothness/variability of spatial property distribution. The developed approach is illustrated with a fluvial reservoir case. The resulting probabilistic production forecasts are described by uncertainty envelopes. The paper compares the performance of the models with different combinations of unknown parameters and discusses sensitivity issues.

Supervised injection services : what has been demonstrated? : a systematic literature review

BACKGROUND: Supervised injection services (SISs) have been developed to promote safer drug injection practices, enhance health-related behaviors among people who inject drugs (PWID), and connect PWID with external health and social services. Nevertheless, SISs have also been accused of fostering drug use and drug trafficking. AIMS: To systematically collect and synthesize the currently available evidence regarding SIS-induced benefits and harm. METHODS: A systematic review was performed via the PubMed, Web of Science, and ScienceDirect databases using the keyword algorithm [("SUPERVISED" OR "SAFER") AND ("INJECTION" OR "INJECTING" OR "SHOOTING" OR "CONSUMPTION") AND ("FACILITY" OR "FACILITIES" OR "ROOM" OR "GALLERY" OR "CENTRE" OR "SITE")]. RESULTS: Seventy-five relevant articles were found. All studies converged to find that SISs were efficacious in attracting the most marginalized PWID, promoting safer injection conditions, enhancing access to primary health care, and reducing the overdose frequency. SISs were not found to increase drug injecting, drug trafficking or crime in the surrounding environments. SISs were found to be associated with reduced levels of public drug injections and dropped syringes. Of the articles, 85% originated from Vancouver or Sydney. CONCLUSION: SISs have largely fulfilled their initial objectives without enhancing drug use or drug trafficking. Almost all of the studies found in this review were performed in Canada or Australia, whereas the majority of SISs are located in Europe. The implementation of new SISs in places with high rates of injection drug use and associated harms appears to be supported by evidence.

An Efficient Segmentation Method for Ultrasound Images based on a Semi-supervised Approach and Patch-based Features

Segmenting ultrasound images is a challenging problemwhere standard unsupervised segmentation methods such asthe well-known Chan-Vese method fail. We propose in thispaper an efficient segmentation method for this class ofimages. Our proposed algorithm is based on asemi-supervised approach (user labels) and the use ofimage patches as data features. We also consider thePearson distance between patches, which has been shown tobe robust w.r.t speckle noise present in ultrasoundimages. Our results on phantom and clinical data show avery high similarity agreement with the ground truthprovided by a medical expert.

Semi-supervised remote sensing image classification with cluster kernels

A semisupervised support vector machine is presented for the classification of remote sensing images. The method exploits the wealth of unlabeled samples for regularizing the training kernel representation locally by means of cluster kernels. The method learns a suitable kernel directly from the image and thus avoids assuming a priori signal relations by using a predefined kernel structure. Good results are obtained in image classification examples when few labeled samples are available. The method scales almost linearly with the number of unlabeled samples and provides out-of-sample predictions.

Geomodelling of a fluvial system with semi-supervised support vector regression

Fluvial deposits are a challenge for modelling flow in sub-surface reservoirs. Connectivity and continuity of permeable bodies have a major impact on fluid flow in porous media. Contemporary object-based and multipoint statistics methods face a problem of robust representation of connected structures. An alternative approach to model petrophysical properties is based on machine learning algorithm ? Support Vector Regression (SVR). Semi-supervised SVR is able to establish spatial connectivity taking into account the prior knowledge on natural similarities. SVR as a learning algorithm is robust to noise and captures dependencies from all available data. Semi-supervised SVR applied to a synthetic fluvial reservoir demonstrated robust results, which are well matched to the flow performance

Breast cancer and melanoma cell line identification by FTIR imaging after formalin-fixation and paraffin-embedding.

Over the past few decades, Fourier transform infrared (FTIR) spectroscopy coupled to microscopy has been recognized as an emerging and potentially powerful tool in cancer research and diagnosis. For this purpose, histological analyses performed by pathologists are mostly carried out on biopsied tissue that undergoes the formalin-fixation and paraffin-embedding (FFPE) procedure. This processing method ensures an optimal and permanent preservation of the samples, making FFPE-archived tissue an extremely valuable source for retrospective studies. Nevertheless, as highlighted by previous studies, this fixation procedure significantly changes the principal constituents of cells, resulting in important effects on their infrared (IR) spectrum. Despite the chemical and spectral influence of FFPE processing, some studies demonstrate that FTIR imaging allows precise identification of the different cell types present in biopsied tissue, indicating that the FFPE process preserves spectral differences between distinct cell types. In this study, we investigated whether this is also the case for closely related cell lines. We analyzed spectra from 8 cancerous epithelial cell lines: 4 breast cancer cell lines and 4 melanoma cell lines. For each cell line, we harvested cells at subconfluence and divided them into two sets. We first tested the "original" capability of FTIR imaging to identify these closely related cell lines on cells just dried on BaF2 slides. We then repeated the test after submitting the cells to the FFPE procedure. Our results show that the IR spectra of FFPE processed cancerous cell lines undergo small but significant changes due to the treatment. The spectral modifications were interpreted as a potential decrease in the phospholipid content and protein denaturation, in line with the scientific literature on the topic. Nevertheless, unsupervised analyses showed that spectral proximities and distances between closely related cell lines were mostly, but not entirely, conserved after FFPE processing. Finally, PLS-DA statistical analyses highlighted that closely related cell lines are still successfully identified and efficiently distinguished by FTIR spectroscopy after FFPE treatment. This last result paves the way towards identification and characterization of cellular subtypes on FFPE tissue sections by FTIR imaging, indicating that this analysis technique could become a potential useful tool in cancer research.

Supervised injection services : what has been demonstrated? A systematic literature review

BACKGROUND: Supervised injection services (SISs) have been developed to promote safer drug injection practices, enhance health-related behaviors among people who inject drugs (PWID), and connect PWID with external health and social services. Nevertheless, SISs have also been accused of fostering drug use and drug trafficking. AIMS: To systematically collect and synthesize the currently available evidence regarding SIS-induced benefits and harm. METHODS: A systematic review was performed via the PubMed, Web of Science, and ScienceDirect databases using the keyword algorithm [("supervised" or "safer") and ("injection" or "injecting" or "shooting" or "consumption") and ("facility" or "facilities" or "room" or "gallery" or "centre" or "site")]. RESULTS: Seventy-five relevant articles were found. All studies converged to find that SISs were efficacious in attracting the most marginalized PWID, promoting safer injection conditions, enhancing access to primary health care, and reducing the overdose frequency. SISs were not found to increase drug injecting, drug trafficking or crime in the surrounding environments. SISs were found to be associated with reduced levels of public drug injections and dropped syringes. Of the articles, 85% originated from Vancouver or Sydney. CONCLUSION: SISs have largely fulfilled their initial objectives without enhancing drug use or drug trafficking. Almost all of the studies found in this review were performed in Canada or Australia, whereas the majority of SISs are located in Europe. The implementation of new SISs in places with high rates of injection drug use and associated harms appears to be supported by evidence.

Biological embedding of early-life exposures and disease risk in humans : a role for DNA methylation

BACKGROUND: Following wider acceptance of 'the thrifty phenotype' hypothesis and the convincing evidence that early-life exposures can influence adult health even decades after the exposure, much interest has been placed on the mechanisms through which early-life exposures become biologically embedded. MATERIALS AND METHODS: In this review, we summarize the current literature regarding biological embedding of early-life experiences. To this end, we conducted a literature search to identify studies investigating early-life exposures in relation to DNA methylation changes. In addition, we summarize the challenges faced in investigations of epigenetic effects, stemming from the peculiarities of this emergent and complex field. A proper systematic review and meta-analyses were not feasible given the nature of the evidence. RESULTS: We identified seven studies on early-life socio-economic circumstances, 10 studies on childhood obesity and six studies on early-life nutrition all relating to DNA methylation changes that met the stipulated inclusion criteria. The pool of evidence gathered, albeit small, favours a role of epigenetics and DNA methylation in biological embedding, but replication of findings, multiple comparison corrections, publication bias and causality are concerns remaining to be addressed in future investigations. CONCLUSIONS: Based on these results, we hypothesize that epigenetics, in particular DNA methylation, is a plausible mechanism through which early-life exposures are biologically embedded. This review describes the current status of the field and acts as a stepping stone for future, better designed investigations on how early-life exposures might become biologically embedded through epigenetic effects.

Large-scale clustering through functional embedding

We present a new framework for large-scale data clustering. The main idea is to modify functional dimensionality reduction techniques to directly optimize over discrete labels using stochastic gradient descent. Compared to methods like spectral clustering our approach solves a single optimization problem, rather than an ad-hoc two-stage optimization approach, does not require a matrix inversion, can easily encode prior knowledge in the set of implementable functions, and does not have an ?out-of-sample? problem. Experimental results on both artificial and real-world datasets show the usefulness of our approach.