Prediction of activity type in preschool children using machine learning techniques

Objectives Recent research has shown that machine learning techniques can accurately predict activity classes from accelerometer data in adolescents and adults. The purpose of this study is to develop and test machine learning models for predicting activity type in preschool-aged children. Design Participants completed 12 standardised activity trials (TV, reading, tablet game, quiet play, art, treasure hunt, cleaning up, active game, obstacle course, bicycle riding) over two laboratory visits. Methods Eleven children aged 3–6 years (mean age = 4.8 ± 0.87; 55% girls) completed the activity trials while wearing an ActiGraph GT3X+ accelerometer on the right hip. Activities were categorised into five activity classes: sedentary activities, light activities, moderate to vigorous activities, walking, and running. A standard feed-forward Artificial Neural Network and a Deep Learning Ensemble Network were trained on features in the accelerometer data used in previous investigations (10th, 25th, 50th, 75th and 90th percentiles and the lag-one autocorrelation). Results Overall recognition accuracy for the standard feed forward Artificial Neural Network was 69.7%. Recognition accuracy for sedentary activities, light activities and games, moderate-to-vigorous activities, walking, and running was 82%, 79%, 64%, 36% and 46%, respectively. In comparison, overall recognition accuracy for the Deep Learning Ensemble Network was 82.6%. For sedentary activities, light activities and games, moderate-to-vigorous activities, walking, and running recognition accuracy was 84%, 91%, 79%, 73% and 73%, respectively. Conclusions Ensemble machine learning approaches such as Deep Learning Ensemble Network can accurately predict activity type from accelerometer data in preschool children.

Design of high mobility DPP-based polymer semiconductors for organic thin film transistors

π-Conjugated polymers are the most promising semiconductor materials to enable printed organic thin film transistors (OTFTs) due to their excellent solution processability and mechanical robustness. However, solution-processed polymer semiconductors have shown poor charge transport properties mainly originated from the disordered polymer chain packing in the solid state as compared to the thermally evaporated small molecular organic semiconductors. The low charge carrier mobility, typically < 0.1 cm2 /V.s, of polymer semiconductors poses a challenge for most intended applications such as displays and radio-frequency identification (RFID) tags. Here we present our recent results on the dike topyrrolopyrrole (DPP)-based polymers and demonstrate that when DPP is combined with appropriate electron donating moieties such as thiophene and thienothiophene, very high charge carrier mobility values of ~1 cm2/V.s could be achieved.

Characterization of acyl chain position in unsaturated phosphatidylcholines using differential mobility-mass spectrometry

Glycerophospholipids (GPs) that differ in the relative position of the two fatty acyl chains on the glycerol backbone (i.e., sn-positional isomers) can have distinct physicochemical properties. The unambiguous assignment of acyl chain position to an individual GP represents a significant analytical challenge. Here we describe a workflow where phosphatidylcholines (PCs) are subjected to ESI for characterization by a combination of differential mobility spectrometry and MS (DMS-MS). When infused as a mixture, ions formed from silver adduction of each phospholipid isomer {e.g., [PC (16:0/18:1) + Ag]+ and [PC (18:1/16:0) + Ag]+} are transmitted through the DMS device at discrete compensation voltages. Varying their relative amounts allows facile and unambiguous assignment of the sn-positions of the fatty acyl chains for each isomer. Integration of the well-resolved ion populations provides a rapid method (< 3 min) for relative quantification of these lipid isomers. The DMS-MS results show excellent agreement with established, but time-consuming, enzymatic approaches and also provide superior accuracy to methods that rely on MS alone. The advantages of this DMS-MS method in identification and quantification of GP isomer populations is demonstrated by direct analysis of complex biological extracts without any prior fractionation.

Adsorption and mobility of metals in build-up on road surfaces

The study investigated the adsorption and bioavailability characteristics of traffic generated metals common to urban land uses, in road deposited solids particles. To validate the outcomes derived from the analysis of field samples, adsorption and desorption experiments were undertaken. The analysis of field samples revealed that metals are selectively adsorbed to different charge sites on solids. Zinc, copper, lead and nickel are adsorbed preferentially to oxides of manganese, iron and aluminium. Lead is adsorbed to organic matter through chemisorption. Cadmium and chromium form weak bonding through cation exchange with most of the particle sizes. Adsorption and desorption experiments revealed that at high metal concentrations, chromium, copper and lead form relatively strong bonds with solids particles while zinc is adsorbed through cation exchange with high likelihood of being released back into solution. Outcomes from this study provide specific guidance for the removal of metals from stormwater based on solids removal.

Better health outcomes for people with physical disabilities in China through audits of mobility related access : a research and implementation need

China has 85 million people with disabilities, 30% of whom have a physical disability(1). Up to 2006, overall disability rates increased by 0.5% per year, more for males and in rural areas, and rates of physical disability increased by 11.2% per year(2). With population ageing the proportion of people with disability will increase even faster. In May 2014 the 67th World Health Assembly adopted a resolution endorsing the WHO Global Disability Action Plan 2014–2021. One of its three objectives is “to remove barriers and improve access to health services and programmes”. Access to transport contributes to positive health outcomes both directly and indirectly (e.g. access to economic opportunities, which is associated with better health)(3). However, once people with physical disabilities leave their dwellings they are confronted with physical barriers to their mobility, ranging from the condition/provision of paths to the cost/availability of transport and access to buildings. In addition, their mobility restrictions increase their vulnerability as road users, exposing them to a higher risk of injury through road crashes. QUT's School of Public Health and Social Work (PHSW) and and Centre for Accident Research and Road Safety-Queensland (CARRS-Q) CARRS-Q have been collaborating on development of a combined disability audit and road safety access tool that can identify transport barriers and safety issues along the routes taken by people with disabilities, to enable prioritisation of actions to address these issues. There are also spin-off benefits for other road users from addressing the rising toll of disability through road crashes in China(4). The tool has undergone initial proof-of-concept testing in India and Viet Nam, and is currently being assessed in Cambodia and Laos. Given the rapid development of China, increases in rates of physical disability and the impacts of an ageing population, it is proposed to establish collaborative research through the Australia-China Centre for Public Health to (1) tailor the combined road safety audit and disability access tool for use in China; (2) evaluate its use on a sample of routes; (3) develop plans for changes to the routes in consultation with local authorities; (4) evaluate the effectiveness of implemented changes in terms of access and health. 1. Zheng, Q, et al, 2014. Health and Quality of Life Outcomes, 12:25. 2. Zheng, X, et al, 2011. Bull World Health Org, 89:788–797. 3. Götschi, T & Kahlmeier, S, 2011. Integrated Transport, Health, and Sustainability Assessment (INTHESA): Final Report. Institute of Social and Preventive Medicine, University of Zurich. 4. Lin, T, et al, 2013. J Public Health, 35:541–547.

The limits to public service : rural communities, professional families, and work mobility

Australia faces an ongoing challenge recruiting professionals to staff essential human services in rural and remote communities. This paper identifies the private limits to the implicit service contract between professions and such client populations. These become evident in how private solutions to competing priorities within professional families inform their selective mobility and thus create the public problem for such communities. The paper reports on a survey of doctors, nurses, teachers and police with responsibility for school-aged children in Queensland that plumbed the strength of neoliberal values in their educational strategy and their commitment to the public good in career decisions. The quantitative analysis suggested that neoliberal values are not necessarily opposed to a commitment to the public good. However, the qualitative analysis of responses to hypothetical career opportunities in rural and remote communities drew out the multiple intertwined spatial and temporal limits to such public service, highlighting the priority given to educational strategy in these families’ deliberations. This private/public nexus poses a policy problem on multiple institutional fronts.

The BAD project: data mining, database and prediction of protein adsorption on surfaces

Protein adsorption at solid-liquid interfaces is critical to many applications, including biomaterials, protein microarrays and lab-on-a-chip devices. Despite this general interest, and a large amount of research in the last half a century, protein adsorption cannot be predicted with an engineering level, design-orientated accuracy. Here we describe a Biomolecular Adsorption Database (BAD), freely available online, which archives the published protein adsorption data. Piecewise linear regression with breakpoint applied to the data in the BAD suggests that the input variables to protein adsorption, i.e., protein concentration in solution; protein descriptors derived from primary structure (number of residues, global protein hydrophobicity and range of amino acid hydrophobicity, isoelectric point); surface descriptors (contact angle); and fluid environment descriptors (pH, ionic strength), correlate well with the output variable-the protein concentration on the surface. Furthermore, neural network analysis revealed that the size of the BAD makes it sufficiently representative, with a neural network-based predictive error of 5% or less. Interestingly, a consistently better fit is obtained if the BAD is divided in two separate sub-sets representing protein adsorption on hydrophilic and hydrophobic surfaces, respectively. Based on these findings, selected entries from the BAD have been used to construct neural network-based estimation routines, which predict the amount of adsorbed protein, the thickness of the adsorbed layer and the surface tension of the protein-covered surface. While the BAD is of general interest, the prediction of the thickness and the surface tension of the protein-covered layers are of particular relevance to the design of microfluidics devices.

Time-independent charge carrier mobility in a model polymer:fullerene organic solar cell

The technique of photo-CELIV (charge extraction by linearly increasing voltage) is one of the more straightforward and popular approaches to measure the faster carrier mobility in measurement geometries that are relevant for operational solar cells and other optoelectronic devices. It has been used to demonstrate a time-dependent photocarrier mobility in pristine polymers, attributed to energetic relaxation within the density of states. Conversely, in solar cell blends, the presence or absence of such energetic relaxation on transport timescales remains under debate. We developed a complete numerical model and performed photo-CELIV experiments on the model high efficiency organic solar cell blend poly[3,6-dithiophene-2-yl-2,5-di(2-octyldodecyl)-pyrrolo[3,4-c]pyrrole-1,4-dione-alt-naphthalene] (PDPP-TNT):[6,6]-phenyl-C71-butyric-acid-methyl-ester (PC70BM). In the studied solar cells a constant, time-independent mobility on the scale relevant to charge extraction was observed, where thermalisation of photocarriers occurs on time scales much shorter than the transit time. Therefore, photocarrier relaxation effects are insignificant for charge transport in these efficient photovoltaic devices.

Learning the learning rate for prediction with expert advice

Most standard algorithms for prediction with expert advice depend on a parameter called the learning rate. This learning rate needs to be large enough to fit the data well, but small enough to prevent overfitting. For the exponential weights algorithm, a sequence of prior work has established theoretical guarantees for higher and higher data-dependent tunings of the learning rate, which allow for increasingly aggressive learning. But in practice such theoretical tunings often still perform worse (as measured by their regret) than ad hoc tuning with an even higher learning rate. To close the gap between theory and practice we introduce an approach to learn the learning rate. Up to a factor that is at most (poly)logarithmic in the number of experts and the inverse of the learning rate, our method performs as well as if we would know the empirically best learning rate from a large range that includes both conservative small values and values that are much higher than those for which formal guarantees were previously available. Our method employs a grid of learning rates, yet runs in linear time regardless of the size of the grid.