Texture Analysis with MTEX – Free and Open Source Software Toolbox
The MATLAB™ toolbox MTEX provides a unique way to represent, analyse and interpret crystallographic preferred orientation, i.e. texture, based on integral (“pole figure”) or individual orientation (“EBSD”) measurements. In particular, MTEX comprises functions to import, analyse and visualize diffraction pole figure data as well as EBSD data, to estimate an orientation density function from either kind of data, to compute texture characteristics, to model orientation density functions in terms of model functions or Fourier coefficients, to simulate pole figure or EBSD data, to create publication ready plots, to write scripts for multiple use, and others. Thus MTEX is a versatile free and open-source software toolbox for texture analysis and modeling.
On the fate of gas accreting at a low rate on to a black hole
Gas supplied conservatively to a black hole at rates well below the Eddington rate may not be able to radiate effectively and the net energy flux, including the energy transported by the viscous torque, is likely to be close to zero at all radii. This has the consequence that the gas accretes with positive energy so that it may escape. Accordingly, we propose that only a small fraction of the gas supplied actually falls on to the black hole, and that the binding energy it releases is transported radially outward by the torque so as to drive away the remainder in the form of a wind. This is a generalization of and an alternative to an `ADAF' solution. Some observational implications and possible ways to distinguish these two types of flow are briefly discussed.
Instance-Aware Semantic Segmentation via Multi-task Network Cascades
Semantic segmentation research has recently witnessed rapid progress, but many leading methods are unable to identify object instances. In this paper, we present Multitask Network Cascades for instance-aware semantic segmentation. Our model consists of three networks, respectively differentiating instances, estimating masks, and categorizing objects. These networks form a cascaded structure, and are designed to share their convolutional features. We develop an algorithm for the nontrivial end-to-end training of this causal, cascaded structure. Our solution is a clean, single-step training framework and can be generalized to cascades that have more stages. We demonstrate state-of-the-art instance-aware semantic segmentation accuracy on PASCAL VOC. Meanwhile, our method takes only 360ms testing an image using VGG-16, which is two orders of magnitude faster than previous systems for this challenging problem. As a by product, our method also achieves compelling object detection results which surpass the competitive Fast/Faster R-CNN systems. The method described in this paper is the foundation of our submissions to the MS COCO 2015 segmentation competition, where we won the 1st place.
Functional brain abnormalities in young adults at genetic risk for late-onset Alzheimer's dementia
Fluorodeoxyglucose positron emission tomography (PET) studies have found that patients with Alzheimer's dementia (AD) have abnormally low rates of cerebral glucose metabolism in posterior cingulate, parietal, temporal, and prefrontal cortex. We previously found that cognitively normal, late-middle-aged carriers of the apolipoprotein E ε4 allele, a common susceptibility gene for late-onset Alzheimer's dementia, have abnormally low rates of glucose metabolism in the same brain regions as patients with probable AD. We now consider whether ε4 carriers have these regional brain abnormalities as relatively young adults. Apolipoprotein E genotypes were established in normal volunteers 20–39 years of age. Clinical ratings, neuropsychological tests, magnetic resonance imaging, and PET were performed in 12 ε4 heterozygotes, all with the ε3/ε4 genotype, and 15 noncarriers of the ε4 allele, 12 of whom were individually matched for sex, age, and educational level. An automated algorithm was used to generate an aggregate surface-projection map that compared regional PET measurements in the two groups. The young adult ε4 carriers and noncarriers did not differ significantly in their sex, age, educational level, clinical ratings, or neuropsychological test scores. Like previously studied patients with probable AD and late-middle-aged ε4 carriers, the young ε4 carriers had abnormally low rates of glucose metabolism bilaterally in the posterior cingulate, parietal, temporal, and prefrontal cortex. Carriers of a common Alzheimer's susceptibility gene have functional brain abnormalities in young adulthood, several decades before the possible onset of dementia.
Fast Global Oscillations in Networks of Integrate-and-Fire Neurons with Low Firing Rates
We study analytically the dynamics of a network of sparsely connected inhibitory integrate-and-fire neurons in a regime where individual neurons emit spikes irregularly and at a low rate. In the limit when the number of neurons N , the network exhibits a sharp transition between a stationary and an oscillatory global activity regime where neurons are weakly synchronized. The activity becomes oscillatory when the inhibitory feedback is strong enough. The period of the global oscillation is found to be mainly controlled by synaptic times but depends also on the characteristics of the external input. In large but finite networks, the analysis shows that global oscillations of finite coherence time generically exist both above and below the critical inhibition threshold. Their characteristics are determined as functions of systems parameters in these two different regimes. The results are found to be in good agreement with numerical simulations.
Comparative bioremediation of soils contaminated with diesel oil by natural attenuation, biostimulation and bioaugmentation.
Bioremediation of diesel oil in soil can occur by natural attenuation, or treated by biostimulation or bioaugmentation. In this study we evaluated all three technologies on the degradation of total petroleum hydrocarbons (TPH) in soil. In addition, the number of diesel-degrading microorganisms present and microbial activity as indexed by the dehydrogenase assay were monitored. Soils contaminated with diesel oil in the field were collected from Long Beach, California, USA and Hong Kong, China. After 12 weeks of incubation, all three treatments showed differing effects on the degradation of light (C12-C23) and heavy (C23-C40) fractions of TPH in the soil samples. Bioaugmentation of the Long Beach soil showed the greatest degradation in the light (72.7%) and heavy (75.2%) fractions of TPH. Natural attenuation was more effective than biostimulation (addition of nutrients), most notably in the Hong Kong soil. The greatest microbial activity (dehydrogenase activity) was observed with bioaugmentation of the Long Beach soil (3.3-fold) and upon natural attenuation of the Hong Kong sample (4.0-fold). The number of diesel-degrading microorganisms and heterotrophic population was not influenced by the bioremediation treatments. Soil properties and the indigenous soil microbial population affect the degree of biodegradation; hence detailed site specific characterization studies are needed prior to deciding on the proper bioremediation method.
Remediation of soils contaminated with polycyclic aromatic hydrocarbons (PAHs).
Polycyclic aromatic hydrocarbons (PAHs) are carcinogenic micropollutants which are resistant to environmental degradation due to their highly hydrophobic nature. Concerns over their adverse health effects have resulted in extensive studies on the remediation of soils contaminated with PAHs. This paper aims to provide a review of the remediation technologies specifically for PAH-contaminated soils. The technologies discussed here include solvent extraction, bioremediation, phytoremediation, chemical oxidation, photocatalytic degradation, electrokinetic remediation, thermal treatment and integrated remediation technologies. For each of these, the theories are discussed in conjunction with comparative evaluation of studies reported in the specialised literature.
An Overview of Multi-task Learning
As a promising area in machine learning, multi-task learning (MTL) aims to improve the performance of multiple related learning tasks by leveraging useful information among them. In this paper, we give an overview of MTL by first giving a definition of MTL. Then several different settings of MTL are introduced, including multi-task supervised learning, multi-task unsupervised learning, multi-task semi-supervised learning, multi-task active learning, multi-task reinforcement learning, multi-task online learning and multi-task multi-view learning. For each setting, representative MTL models are presented. In order to speed up the learning process, parallel and distributed MTL models are introduced. Many areas, including computer vision, bioinformatics, health informatics, speech, natural language processing, web applications and ubiquitous computing, use MTL to improve the performance of the applications involved and some representative works are reviewed. Finally, recent theoretical analyses for MTL are presented.
Combined measurements of Subglacial Water Pressure and Surface Velocity of Findelengletscher, Switzerland: Conclusions about Drainage System and Sliding Mechanism
During the snow-melt season of 1982, basal water pressure was recorded in 11 bore holes communicating with the subglacial drainage system. In most of these holes the water levels were at approximately the same depth (around 70 m below surface). The large variations of water pressure, such as diurnal variations, were usually similar at different locations and in phase. In two instances of exceptionally high water pressure, however, systematic phase shifts were observed; a wave of high pressure travelled down-glacier with a velocity of approximately 100 m/h. The glacier-surface velocity was measured at four lines of stakes several times daily. The velocity variations correlated with variations in subglacial water pressure. The functional relationship of water pressure and velocity suggests that fluctuating bed separation was responsible for the velocity variations. The empirical functional relationship is compared to that of sliding over a perfectly lubricated sinusoidal bed. On the basis of the measured velocity-pressure relationship, this model predicts a reasonable value of bed roughness but too high a sliding velocity and unstable sliding at too low a water pressure. The main reason for this disagreement is probably the neglect of friction from debris in the sliding model. The measured water pressure was considerably higher than that predicted by the theory of steady flow through straight cylindrical channels near the glacier bed. Possible reasons are considered. The very large disagreement between measured and predicted pressure suggests that no straight cylindrical channels may have existed.
Perspectives on Free and Open Source Software
What is the status of the Free and Open Source Software (F/OSS) revolution? Has the creation of software that can be freely used, modified, and redistributed transformed industry and society, as some predicted, or is this transformation still a work in progress? Perspectives on Free and Open Source Software brings together leading analysts and researchers to address this question, examining specific aspects of F/OSS in a way that is both scientifically rigorous and highly relevant to real-life managerial and technical concerns. The book analyzes a number of key topics: the motivation behind F/OSS—why highly skilled software developers devote large amounts of time to the creation of "free" products and services; the objective, empirically grounded evaluation of software—necessary to counter what one chapter author calls the "steamroller" of F/OSS hype; the software engineering processes and tools used in specific projects, including Apache, GNOME, and Mozilla; the economic and business models that reflect the changing relationships between users and firms, technical communities and firms, and between competitors; and legal, cultural, and social issues, including one contribution that suggests parallels between "open code" and "open society" and another that points to the need for understanding the movement's social causes and consequences.
Recent Findings on the Phytoremediation of Soils Contaminated with Environmentally Toxic Heavy Metals and Metalloids Such as Zinc, Cadmium, Lead, and Arsenic
Due to their immutable nature, metals are a group of pollutants of much concern. As a result of human activities such as mining and smelting of metalliferous ores, electroplating, gas exhaust, energy and fuel production, fertilizer and pesticide application, etc., metal pollution has become one of the most serious environmental problems today. Phytoremediation, an emerging cost-effective, non-intrusive, and aesthetically pleasing technology, that uses the remarkable ability of plants to concentrate elements and compounds from the environment and to metabolize various molecules in their tissues, appears very promising for the removal of pollutants from the environment. Within this field of phytoremediation, the utilization of plants to transport and concentrate metals from the soil into the harvestable parts of roots and above-ground shoots, i.e., phytoextraction, may be, at present, approaching commercialization. Improvement of the capacity of plants to tolerate and accumulate metals by genetic engineering should open up new possibilities for phytoremediation. The lack of understanding pertaining to metal uptake and translocation mechanisms, enhancement amendments, and external effects of phytoremediation is hindering its full scale application. Due to its great potential as a viable alternative to traditional contaminated land remediation methods, phytoremediation is currently an exciting area of active research.
Clustered Multi-Task Learning: A Convex Formulation
In multi-task learning several related tasks are considered simultaneously, with the hope that by an appropriate sharing of information across tasks, each task may benefit from the others. In the context of learning linear functions for supervised classification or regression, this can be achieved by including a priori information about the weight vectors associated with the tasks, and how they are expected to be related to each other. In this paper, we assume that tasks are clustered into groups, which are unknown beforehand, and that tasks within a group have similar weight vectors. We design a new spectral norm that encodes this a priori assumption, without the prior knowledge of the partition of tasks into groups, resulting in a new convex optimization formulation for multi-task learning. We show in simulations on synthetic examples and on the IEDB MHC-I binding dataset, that our approach outperforms well-known convex methods for multi-task learning, as well as related non-convex methods dedicated to the same problem.
Melting of a peridotite nodule at high pressures and high water pressures
Melting of a natural peridotite (spinel-bearing lherzolite) which occurs as a nodule in the tuff of Salt Lake, Hawaii, has been studied at pressures between 1 atm and 50 kb under anhydrous conditions and at pressures between 20 and 60 kb under hydrous conditions with the tetrahedral-anvil type of high-pressure apparatus. Under anhydrous conditions the lherzolite begins to melt near the liquidus of some olivine tholeiites. Garnet is stable near the solidus at pressures higher than at least 30 kb. Under hydrous conditions, when sealed capsules are used, the solidus of the lherzolite is at about 1000°C at 26 kb and about 1150°C at 60 kb. It is 400–700°C lower than the solidus under anhydrous conditions. When unsealed capsules are used, the solidus is raised by 200–400°C from the solidus determined by using sealed capsules. From the present experiments it appears that under anhydrous conditions magmas of olivine tholeiite composition can be formed from lherzolite, but those of quartz-tholeiite composition cannot be formed by partial melting, at least in the pressure range 10–30 kb. Quartz-tholeiite magma, however, can be formed within a much larger pressure range under hydrous conditions. The solidus under hydrous conditions (water pressure is equal to total pressure) would give a possible lowest temperature of beginning of melting of the upper mantle. It is also suggested that the partial melting of the hydrous upper mantle may play an important part in the formation of the low-velocity zone.
Tree-Guided Group Lasso for Multi-Task Regression with Structured Sparsity
We consider the problem of learning a sparse multi-task regression, where the structure in the outputs can be represented as a tree with leaf nodes as outputs and internal nodes as clusters of the outputs at multiple granularity. Our goal is to recover the common set of relevant inputs for each output cluster. Assuming that the tree structure is available as prior knowledge, we formulate this problem as a new multi-task regularized regression called tree-guided group lasso. Our structured regularization is based on a group-lasso penalty, where groups are defined with respect to the tree structure. We describe a systematic weighting scheme for the groups in the penalty such that each output variable is penalized in a balanced manner even if the groups overlap. We present an efficient optimization method that can handle a large-scale problem. Using simulated and yeast datasets, we demonstrate that our method shows a superior performance in terms of both prediction errors and recovery of true sparsity patterns compared to other methods for multi-task learning.
Learning Task Grouping and Overlap in Multi-task Learning
In the paradigm of multi-task learning, multiple related prediction tasks are learned jointly, sharing information across the tasks. We propose a framework for multi-task learning that enables one to selectively share the information across the tasks. We assume that each task parameter vector is a linear combination of a finite number of underlying basis tasks. The coefficients of the linear combination are sparse in nature and the overlap in the sparsity patterns of two tasks controls the amount of sharing across these. Our model is based on the assumption that task parameters within a group lie in a low dimensional subspace but allows the tasks in different groups to overlap with each other in one or more bases. Experimental results on four datasets show that our approach outperforms competing methods.
A Convex Formulation for Learning Task Relationships in Multi-Task Learning
Multi-task learning is a learning paradigm which seeks to improve the generalization performance of a learning task with the help of some other related tasks. In this paper, we propose a regularization formulation for learning the relationships between tasks in multi-task learning. This formulation can be viewed as a novel generalization of the regularization framework for single-task learning. Besides modeling positive task correlation, our method, called multi-task relationship learning (MTRL), can also describe negative task correlation and identify outlier tasks based on the same underlying principle. Under this regularization framework, the objective function of MTRL is convex. For efficiency, we use an alternating method to learn the optimal model parameters for each task as well as the relationships between tasks. We study MTRL in the symmetric multi-task learning setting and then generalize it to the asymmetric setting as well. We also study the relationships between MTRL and some existing multi-task learning methods. Experiments conducted on a toy problem as well as several benchmark data sets demonstrate the effectiveness of MTRL.
Robust Visual Tracking via Structured Multi-Task Sparse Learning
In this paper, we formulate object tracking in a particle filter framework as a structured multi-task sparse learning problem, which we denote as Structured Multi-Task Tracking (S-MTT). Since we model particles as linear combinations of dictionary templates that are updated dynamically, learning the representation of each particle is considered a single task in Multi-Task Tracking (MTT). By employing popular sparsity-inducing $$\ell _{p,q}$$ mixed norms $$(\text{ specifically} p\in \{2,\infty \}$$ and $$q=1),$$ we regularize the representation problem to enforce joint sparsity and learn the particle representations together. As compared to previous methods that handle particles independently, our results demonstrate that mining the interdependencies between particles improves tracking performance and overall computational complexity. Interestingly, we show that the popular $$L_1$$ tracker (Mei and Ling, IEEE Trans Pattern Anal Mach Intel 33(11):2259–2272, 2011) is a special case of our MTT formulation (denoted as the $$L_{11}$$ tracker) when $$p=q=1.$$ Under the MTT framework, some of the tasks (particle representations) are often more closely related and more likely to share common relevant covariates than other tasks. Therefore, we extend the MTT framework to take into account pairwise structural correlations between particles (e.g. spatial smoothness of representation) and denote the novel framework as S-MTT. The problem of learning the regularized sparse representation in MTT and S-MTT can be solved efficiently using an Accelerated Proximal Gradient (APG) method that yields a sequence of closed form updates. As such, S-MTT and MTT are computationally attractive. We test our proposed approach on challenging sequences involving heavy occlusion, drastic illumination changes, and large pose variations. Experimental results show that S-MTT is much better than MTT, and both methods consistently outperform state-of-the-art trackers.
Deep multi-task learning with low level tasks supervised at lower layers
In all previous work on deep multi-task learning we are aware of, all task supervisions are on the same (outermost) layer. We present a multi-task learning architecture with deep bi-directional RNNs, where different tasks supervision can happen at different layers. We present experiments in syntactic chunking and CCG supertagging, coupled with the additional task of POS-tagging. We show that it is consistently better to have POS supervision at the innermost rather than the outermost layer. We argue that this is because “lowlevel” tasks are better kept at the lower layers, enabling the higher-level tasks to make use of the shared representation of the lower-level tasks. Finally, we also show how this architecture can be used for domain adaptation.
IEEE 802.15.4 low rate - wireless personal area network coexistence issues
IEEE 802.15.4 is a proposed standard addressing the needs of low-rate wireless personal area networks or LR-WPAN with a focus on enabling wireless sensor networks. The standard is characterized by maintaining a high level of simplicity, allowing for low cost and low power implementations. Its operational frequency band includes the 2.4 GHz industrial, scientific and medical band providing nearly worldwide availability; additionally, this band is also used by other IEEE 802 wireless standards. Coexistence among diverse collocated devices in the 2.4 GHz band is an important issue in order to ensure that each wireless service maintains its desired performance requirements. This paper presents a brief technical introduction of the IEEE 802.15.4 standard and analyzes the coexistence impact of an IEEE 802.15.4 network on the IEEE 802.11b devices.
Laboratory Analysis of Mudflow Properties
A rotational viscometer has been designed for laboratory measurements of the rheological properties of natural mudflow deposits in Colorado. The mudflow matrices comprised of silt and clay particles are sheared under temperature-controlled conditions at volumetric sediment concentrations ranging from 0.10 – 0.45. This study stresses the importance of conducting rheological measurements at low rates of shear because: 1) Those are the conditions found in natural channels; and 2) they avoid the slippage problems observed at large sediment concentrations. At low rates of shear, the Bingham model is fitted to the measured rheograms, and both the viscosity and yield stress increase exponentially with the sediment concentration of the fluid matrix. Both the yield stress and the viscosity increase by three orders of magnitude as the volumetric concentration of sediments in the fluid matrix changes from 10% to 40%. The addition of sand particles does not significantly alter the rheological properties of the matrix unless the volumetric concentration of sands exceeds 0.20.
neural network differential equation deep learning convolutional neural network convolutional neural software development deep neural network feature selection open source deep neural optical network learning approach geographic information system distributed generation geographic information stochastic differential equation power allocation open source software stochastic differential deep convolutional neural source software deep convolutional named entity recognition loss function human action space telescope personal area network mental health feature learning wireless personal area elastic optical multi-task learning personal area elastic optical network network reconfiguration wireless personal learning problem hubble space telescope learning task software developer high rate water pressure hubble space doubly stochastic open source tool source software development free and open capacity allocation low rate source software project group lasso path protection sparse learning backup path rate wireles open source gi spare capacity optimal reconfiguration source gi multi-task deep backward doubly stochastic doubly stochastic differential backward doubly big bang-big crunch low rate wireles survivable routing multi-task deep learning guidance sensor bang-big crunch big bang-big multi-task learning approach source software tool deep multi-task deep multi-task learning pose-invariant face high water pressure fine guidance shared backup fine guidance sensor multi-task convolutional neural multi-task learning model source gis software shared backup path multi-task feature bang-big crunch algorithm backup path protection approximating optimal crunch algorithm multi-task learning method software movement multi-task network spare capacity allocation multi-task convolutional multi-task cnn multi-task sparse multi-task feature learning multi-task regression multi-task allocation feature learning algorithm multiple related task multi-task multi-view multi-task learning algorithm soils contaminated