Topic model kernel : an empirical study towards probabilistically reduced features for classification

Probabilistic topic models have become a standard in modern machine learning with wide applications in organizing and summarizing ‘documents’ in high-dimensional data such as images, videos, texts, gene expression data, and so on. Representing data by dimensional reduction of mixture proportion extracted from topic models is not only richer in semantics than bag-of-word interpretation, but also more informative for classification tasks. This paper describes the Topic Model Kernel (TMK), a high dimensional mapping for Support Vector Machine classification of data generated from probabilistic topic models. The applicability of our proposed kernel is demonstrated in several classification tasks from real world datasets. We outperform existing kernels on the distributional features and give the comparative results on non-probabilistic data types.

A study of seed viability following consumption by birds

This project aimed to quantify the potential for wild birds to spread genetically modified seeds from their area of planting to adjacent areas, through ingestion and transport in the gut (endozoochorous transmission). We addressed this by assessing whether seed viability was affected by passage through the avian gut. We sought to address this question by testing the effects of ingestion of a number of key agricultural crops by three common wild bird pest species.

Examining the viability of windows impact/update method for delay analysis in construction disputes

Nowadays, construction delay disputes often end up on the arbitration route where the delay experts appointed by the parties advise the tribunal on the extension of times entitlements of the parties. For this purpose, the identification and quantification of concurrent and pacing delays are integral aspects of resolving these disputes using a proper delay analysis methodology. The aim of the study is therefore, threefold. Firstly, the available literature on the concurrent and pacing delays are analyzed in detail to establish the principles for the evaluation of the concurrency and pacing delays. Secondly, a robust delay analysis methodology called ‘windows impact/update method’ is explained often used by the experts for the effective quantification of concurrent and pacing delays. This methodology is an improved version of time impact analysis and normal windows analysis. For better demonstration, the explanation of the methodology is facilitated with the help of a typical case study analysis. Finally, the principles of concurrency and pacing, as explained in the literature review, are promptly applied to the case study results to show the applicability of the analysis method on any types of delay disputes. The study shows the effectiveness of the windows impact/update method for the quantification of the concurrent and pacing delays.

Effects of rising temperature on the viability of an important sea turtle rookery

A warming world poses challenges for species with temperature-dependent sex determination, including sea turtles, for which warmer incubation temperatures produce female hatchlings. We combined in situ sand temperature measurements with air temperature records since 1850 and predicted warming scenarios from the Intergovernmental Panel on Climate Change to derive 250-year time series of incubation temperatures, hatchling sex ratios, and operational sex ratios for one of the largest sea turtles rookeries globally (Cape Verde Islands, Atlantic). We estimate that light-coloured beaches currently produce 70.10% females whereas dark-coloured beaches produce 93.46% females. Despite increasingly female skewed sex ratios, entire feminization of this population is not imminent. Rising temperatures increase the number of breeding females and hence the natural rate of population growth. Predicting climate warming impacts across hatchlings, male-female breeding ratios and nesting numbers provides a holistic approach to assessing the conservation concerns for sea turtles in a warming world. © 2014 Macmillan Publishers Limited.

The use of kernel density estimation to examine associations between neighborhood destination intensity and walking and physical activity

BACKGROUND: Local destinations have previously been shown to be associated with higher levels of both physical activity and walking, but little is known about how the distribution of destinations is related to activity. Kernel density estimation is a spatial analysis technique that accounts for the location of features relative to each other. Using kernel density estimation, this study sought to investigate whether individuals who live near destinations (shops and service facilities) that are more intensely distributed rather than dispersed: 1) have higher odds of being sufficiently active; 2) engage in more frequent walking for transport and recreation. METHODS: The sample consisted of 2349 residents of 50 urban areas in metropolitan Melbourne, Australia. Destinations within these areas were geocoded and kernel density estimates of destination intensity were created using kernels of 400m (meters), 800m and 1200m. Using multilevel logistic regression, the association between destination intensity (classified in quintiles Q1(least)-Q5(most)) and likelihood of: 1) being sufficiently active (compared to insufficiently active); 2) walking≥4/week (at least 4 times per week, compared to walking less), was estimated in models that were adjusted for potential confounders. RESULTS: For all kernel distances, there was a significantly greater likelihood of walking≥4/week, among respondents living in areas of greatest destinations intensity compared to areas with least destination intensity: 400m (Q4 OR 1.41 95%CI 1.02-1.96; Q5 OR 1.49 95%CI 1.06-2.09), 800m (Q4 OR 1.55, 95%CI 1.09-2.21; Q5, OR 1.71, 95%CI 1.18-2.48) and 1200m (Q4, OR 1.7, 95%CI 1.18-2.45; Q5, OR 1.86 95%CI 1.28-2.71). There was also evidence of associations between destination intensity and sufficient physical activity, however these associations were markedly attenuated when walking was included in the models. CONCLUSIONS: This study, conducted within urban Melbourne, found that those who lived in areas of greater destination intensity walked more frequently, and showed higher odds of being sufficiently physically active-an effect that was largely explained by levels of walking. The results suggest that increasing the intensity of destinations in areas where they are more dispersed; and or planning neighborhoods with greater destination intensity, may increase residents' likelihood of being sufficiently active for health.

Viability of tallow inclusion in Atlantic salmon diet, as assessed by an on-farm grow out trial

The results of a recent increase in research interest directed at the inclusion of tallow in fish feed formulations are suggesting tallow is viable as a potential substitute for other alternative lipid sources such as poultry by-product oil. Although strong growth performance data has been shown, reservations still exist regarding reduced digestibility and the potential impacts this could have on performance over the duration of a grow-out period in low temperature conditions. Also little information is yet available on the potential effect of dietary tallow inclusion on final product quality. A large scale farm based study testing the inclusion of tallow at 40% inclusion, partially replacing poultry by-product oil, in commercial diets of Atlantic salmon over a winter grow-out period in southern Tasmania, Australia was conducted. Tallow inclusion had no impact on growth performance or nutrient digestibility. Tallow resulted in a slight improvement in fillet quality exhibiting a significant reduction in n - 6 PUFA and the n6:n3 ratio, and an increased n - 3LC-PUFA tissue deposition. Consumers were unable to display any preference in liking between 3 salmon products (cold smoked, hot smoked, and cooked) as a result of tallow inclusion. This study demonstrates the viability of partial inclusion of tallow in Atlantic salmon diets over a winter grow-out period. Statement of relevance: Improved knowledge of alternative dietary energy sources (oils and fats) to be used in aquafeed, (replacing the increasingly expensive, and diminishingly available, fish oil) is a key area of research towards improved environmental sustainability and economic viability of the aquaculture sector. Following a promising laboratory based, research scale, in vivo trial aimed at assessing the viability of tallow in salmon feed, a larger and longer duration farm-based trial was implemented to validate initial findings. Consumer test of final products (fresh-cooked, hot smoked and cold smoked fillets) showed no modification of sensorial attributes. Tallow is hereto shown to be a highly viable alternative oil for the salmon aquafeed industry.

Topic model kernel classification with probabilistically reduced features

Probabilistic topic models have become a standard in modern machine learning to deal with a wide range of applications. Representing data by dimensional reduction of mixture proportion extracted from topic models is not only richer in semantics interpretation, but could also be informative for classification tasks. In this paper, we describe the Topic Model Kernel (TMK), a topicbased kernel for Support Vector Machine classification on data being processed by probabilistic topic models. The applicability of our proposed kernel is demonstrated in several classification tasks with real world datasets. TMK outperforms existing kernels on the distributional features and give comparative results on nonprobabilistic data types.

Derivation of dynamic material properties of steel-fibre-reinforced concrete using kernel regression

The reliable and efficient design of steel-fibre-reinforced concrete (SFRC) structures requires clear knowledge of material properties. Since the locations and orientations of aggregates and fibres in concrete are intrinsically random, testing results from different specimens vary, and it needs hundreds or even thousands of specimens and tests to derive the unbiased statistical distributions of material properties by using traditional statistical techniques. Therefore, few statistical studies on the SFRC material properties can be found in literature. In this study, high-rate impact test results on SFRC using split Hopkinson pressure bar are further analysed. The influences of different strain rates and various volume fractions of fibres on compressive strength of SFRC specimens under dynamic loadings will be quantified, by using kernel regression, a kernel-based nonparametric statistical method. Several kernel estimators and functions will be compared. This technique allows one to derive an unbiased statistical estimation from limited testing data. Therefore it is especially useful when the testing data is limited.

Predicting climate warming effects on green turtle hatchling viability and dispersal performance

Ectotherms are taxa considered highly sensitive to rapid climate warming. This is because body temperature profoundly governs their performance, fitness and life history. Yet, while several modelling approaches currently predict thermal effects on some aspects of life history and demography, they do not consider how temperature simultaneously affects developmental success and offspring phenotypic performance, two additional key attributes that are needed to comprehensively understand species responses to climate warming. Here, we developed a stepwise, individual-level modelling approach linking biophysical and developmental models with empirically derived performance functions to predict the effects of temperature-induced changes to offspring viability, phenotype and performance, using green sea turtle hatchlings as an ectotherm model. Climate warming is expected to particularly threaten sea turtles, as their life-history traits may preclude them from rapid adaptation. Under conservative and extreme warming, our model predicted large effects on performance attributes key to dispersal, as well as a reduction in offspring viability. Forecast sand temperatures produced smaller, weaker hatchlings, which were up to 40% slower than at present, albeit with increased energy stores. Conversely, increases in sea surface temperatures aided swimming performance. Our exploratory study points to the need for further development of integrative individual-based modelling frameworks to better understand the complex outcomes of climate change for ectotherm species. Such advances could better serve ecologists to highlight the most vulnerable species and populations, encouraging prioritization of conservation effort to the most threatened systems.

Using kernel density estimation to understand the influence of neighbourhood destinations on BMI

OBJECTIVES: Little is known about how the distribution of destinations in the local neighbourhood is related to body mass index (BMI). Kernel density estimation (KDE) is a spatial analysis technique that accounts for the location of features relative to each other. Using KDE, this study investigated whether individuals living near destinations (shops and service facilities) that are more intensely distributed rather than dispersed, have lower BMIs.

STUDY DESIGN AND SETTING: A cross-sectional study of 2349 residents of 50 urban areas in metropolitan Melbourne, Australia.

METHODS: Destinations were geocoded, and kernel density estimates of destination intensity were created using kernels of 400, 800 and 1200 m. Using multilevel linear regression, the association between destination intensity (classified in quintiles Q1(least)-Q5(most)) and BMI was estimated in models that adjusted for the following confounders: age, sex, country of birth, education, dominant household occupation, household type, disability/injury and area disadvantage. Separate models included a physical activity variable.

RESULTS: For kernels of 800 and 1200 m, there was an inverse relationship between BMI and more intensely distributed destinations (compared to areas with least destination intensity). Effects were significant at 1200 m: Q4, β -0.86, 95% CI -1.58 to -0.13, p=0.022; Q5, β -1.03 95% CI -1.65 to -0.41, p=0.001. Inclusion of physical activity in the models attenuated effects, although effects remained marginally significant for Q5 at 1200 m: β -0.77 95% CI -1.52, -0.02, p=0.045.

CONCLUSIONS: This study conducted within urban Melbourne, Australia, found that participants living in areas of greater destination intensity within 1200 m of home had lower BMIs. Effects were partly explained by physical activity. The results suggest that increasing the intensity of destination distribution could reduce BMI levels by encouraging higher levels of physical activity.

Inhibition of HDAC3- and HDAC6-promoted survivin expression plays an important role in SAHA-induced autophagy and viability reduction in breast cancer cells

SAHA is a class I HDAC/HDAC6 co-inhibitor and an autophagy inducer currently undergoing clinical investigations in breast cancer patients. However, the molecular mechanism of action of SAHA in breast cancer cells remains unclear. In this study, we found that SAHA is equally effective in targeting cells of different breast cancer subtypes and tamoxifen sensitivity. Importantly, we found that down-regulation of survivin plays an important role in SAHA-induced autophagy and cell viability reduction in human breast cancer cells. SAHA decreased survivin and XIAP gene transcription, induced survivin protein acetylation and early nuclear translocation in MCF7 and MDA-MB-231 breast cancer cells. It also reduced survivin and XIAP protein stability in part through modulating the expression and activation of the 26S proteasome and heat-shock protein 90. Interestingly, targeting HDAC3 and HDAC6, but not other HDAC isoforms, by siRNA/pharmacological inhibitors mimicked the effects of SAHA in modulating the acetylation, expression, and nuclear translocation of survivin and induced autophagy in MCF7 and MDA-MB-231 cancer cells. Targeting HDAC3 also mimicked the effect of SAHA in up-regulating the expression and activity of proteasome, which might lead to the reduced protein stability of survivin in breast cancer cells. In conclusion, this study provides new insights into SAHA's molecular mechanism of actions in breast cancer cells. Our findings emphasize the complexity of the regulatory roles in different HDAC isoforms and potentially assist in predicting the mechanism of novel HDAC inhibitors in targeted or combinational therapies in the future.

Using climatic suitability thresholds to identify past, present and future population viability

Often climatic niche models predict that any change in climatic conditions will impact species abundance or distribution. However, the accuracy of models that just incorporate climatic information to predict future species habitat use is widely debated. Alternatively, environmental conditions may simply need to be above some minimum threshold of climatic suitability, at which point, other factors drive population size. Using the example of nesting sites of loggerhead sea turtles (Caretta caretta) in the Mediterranean (n = 105), we developed climatic niche models to examine whether a climatic suitability threshold could be identified as a climatic indicator in order for large populations of a widespread species to exist. We then assessed the climatic suitability of sites above and below this threshold in the past (∼1900) and future (∼2100). Most large sites that are currently above the climatic threshold were above the threshold in the past and future, particularly when future nesting seasonality shifted to start 1–2 months earlier. Our analyses highlight the importance of future phenological shifts for maintaining suitability. Our results provide a positive outlook for sea turtle conservation, suggesting that climatic conditions may remain suitable in the future at sites that currently support large nesting populations. Our study also provides an alternative way of interpreting the outputs of climatic niche models, by generating a threshold as an index of a minimum climatic suitability required to sustain large populations. This type of approach offers the possibility to benefit from information provided by climate-driven models, while reducing their inherent uncertainties.