Computing and Information Systems - Research Publications
Permanent URI for this collection
Search Results
1 - 10 of 1352
-
ItemNo Preview AvailableModeling of microalgal shear-induced flocculation and sedimentation using a coupled CFD-population balance approach.(Wiley, 2018)In this study, shear-induced flocculation modeling of Chlorella sp. microalgae was conducted by combination of population balance modeling and CFD. The inhomogeneous Multiple Size Group (MUSIG) and the Euler-Euler two fluid models were coupled via Ansys-CFX-15 software package to achieve both fluid and particle dynamics during the flocculation. For the first time, a detailed model was proposed to calculate the collision frequency and breakage rate during the microalgae flocculation by means of the response surface methodology as a tool for optimization. The particle size distribution resulted from the model was in good agreement with that of the jar test experiment. Furthermore, the subsequent sedimentation step was also examined by removing the shear rate in both simulations and experiments. Consequently, variation in the shear rate and its effects on the flocculation behavior, sedimentation rate and recovery efficiency were evaluated. Results indicate that flocculation of Chlorella sp. microalgae under shear rates of 37, 182, and 387 s-1 is a promising method of pre-concentration which guarantees the cost efficiency of the subsequent harvesting process by recovering more than 90% of the biomass.
-
ItemNye tray vs sieve tray: A comparison based on computational fluid dynamics and tray efficiency(Wiley, 2021-10)Nye and sieve trays were hydrodynamically simulated and compared. The simulations were performed in a Eulerian‐Eulerian framework under unsteady (transient) conditions at industrial scale. Conducted on an air‐water system, the simulations included three dimensions and two phases. The velocity distribution across the tray, the height of clear liquid, the froth height, and the pressure drop were investigated and compared with experimental data. Péclet number was calculated using hydrodynamic and geometric parameters. The tray efficiencies were also predicted to further compare the two trays. The results showed that the liquid flow was steadier on the Nye tray rather than the sieve tray, possibly because of the special structure of the liquid and gas inlets for the Nye tray.
-
ItemComputational Fluid Dynamics versus Experiment: An Investigation on Liquid Weeping of Nye Trays(Wiley, 2021-01)The weeping phenomenon was investigated using some experimental tests and a numerical model. The tests were performed within a 1.22‐m‐diameter pilot‐scale column including two chimney trays and two Nye test trays with an air‐water system. The rates of weeping were measured in the Nye trays with two heights of the weir and a hole area of 5 %. Moreover, the weeping rates in the outlet and inlet halves of the Nye tray and the total weeping rate were calculated. In the next step, an Eulerian‐Eulerian computational fluid dynamics (CFD) technique was used. The results show good agreement between the attained CFD findings and the experimental data.
-
ItemExperimental investigation, numerical simulation and RSM modelling of the freezing and thawing of Mozzarella cheese(Elsevier, 2024-01)Freezing can be used to preserve functionality of Mozzarella cheese allowing export to distant markets but limited tools are available for prediction of freezing and thawing times as a function of composition and processing variables. Freezing and thawing processes were experimentally and numerically assessed for six Mozzarella samples, differing significantly in block size and composition. Numerical simulations using an enthalpy method were developed to build a validated and robust model for solving heat and mass transfer equations. A decrease in salt (NaCl) content from 1.34 % w/w to 0.07 % significantly altered the temperature of phase change from ∼–4.5 °C to –3 °C. Simulations showed minimal impact of salt migration on the salt in free moisture content deeper than ∼1–2 centimeters from the surface during freezing, with a slight increase of 8–10 % salt in free moisture at the block center. A response surface methodology (RSM) model was fit to the simulated data providing a useful tool for predicting freezing and thawing times for block sizes and a wider range of operating conditions enabling future process optimization. The RSM model indicated that increased salt content increased freezing time but decreased thawing time.
-
ItemMachine learning for the prediction of proteolysis in Mozzarella and Cheddar cheese(Elsevier, 2024-03)Proteolysis is a complex biochemical event during cheese storage that affects both functionality and quality, yet there are few tools that can accurately predict proteolysis for Mozzarella and Cheddar cheese across a range of parameters and storage conditions. Machine learning models were developed with input features from the literature. A gradient boosting method outperformed random forest and support vector regression methods in predicting proteolysis for both Mozzarella (R2 = 92%) and Cheddar (R2 = 97%) cheese. Storage time was the most important input feature for both cheese types, followed by coagulating enzyme concentration and calcium content for Mozzarella cheese and fat or moisture content for Cheddar cheese. The ability to predict proteolysis could be useful for manufacturers, assisting in inventory management to ensure optimum Mozzarella functionality and Cheddar with a desired taste, flavor and texture; this approach may also be extended to other types of cheese.
-
ItemPriority populations' experiences of isolation, quarantine and distancing for COVID-19: protocol for a longitudinal cohort study (Optimise Study)(BMJ PUBLISHING GROUP, 2024-01)INTRODUCTION: Longitudinal studies can provide timely and accurate information to evaluate and inform COVID-19 control and mitigation strategies and future pandemic preparedness. The Optimise Study is a multidisciplinary research platform established in the Australian state of Victoria in September 2020 to collect epidemiological, social, psychological and behavioural data from priority populations. It aims to understand changing public attitudes, behaviours and experiences of COVID-19 and inform epidemic modelling and support responsive government policy. METHODS AND ANALYSIS: This protocol paper describes the data collection procedures for the Optimise Study, an ongoing longitudinal cohort of ~1000 Victorian adults and their social networks. Participants are recruited using snowball sampling with a set of seeds and two waves of snowball recruitment. Seeds are purposively selected from priority groups, including recent COVID-19 cases and close contacts and people at heightened risk of infection and/or adverse outcomes of COVID-19 infection and/or public health measures. Participants complete a schedule of monthly quantitative surveys and daily diaries for up to 24 months, plus additional surveys annually for up to 48 months. Cohort participants are recruited for qualitative interviews at key time points to enable in-depth exploration of people's lived experiences. Separately, community representatives are invited to participate in community engagement groups, which review and interpret research findings to inform policy and practice recommendations. ETHICS AND DISSEMINATION: The Optimise longitudinal cohort and qualitative interviews are approved by the Alfred Hospital Human Research Ethics Committee (# 333/20). The Optimise Study CEG is approved by the La Trobe University Human Ethics Committee (# HEC20532). All participants provide informed verbal consent to enter the cohort, with additional consent provided prior to any of the sub studies. Study findings will be disseminated through public website (https://optimisecovid.com.au/study-findings/) and through peer-reviewed publications. TRIAL REGISTRATION NUMBER: NCT05323799.
-
ItemEnhancing constraint programming via supervised learning for job shop scheduling(Elsevier, 2024-06-07)Constraint programming (CP) is a powerful technique for solving constraint satisfaction and optimization problems. In CP solvers, the variable ordering strategy used to select which variable to explore first in the solving process has a significant impact on solver effectiveness. To address this issue, we propose a novel variable ordering strategy based on supervised learning, which we evaluate in the context of job shop scheduling problems. Our learning-based methods predict the optimal solution of a problem instance and use the predicted solution to order variables for CP solvers. Unlike traditional variable ordering methods, our methods can learn from the characteristics of each problem instance and customize the variable ordering strategy accordingly, leading to improved solver performance. Our experiments demonstrate that training machine learning models is highly efficient and can achieve high accuracy. Furthermore, our learned variable ordering methods perform competitively compared to four existing methods. Finally, we showcase the benefits of integrating machine learning-based variable ordering methods with conventional domain-based approaches through tie-breaking.
-
ItemBenchmarks for measurement of duplicate detection methods in nucleotide databases(OXFORD UNIV PRESS, 2023-12-18)UNLABELLED: Duplication of information in databases is a major data quality challenge. The presence of duplicates, implying either redundancy or inconsistency, can have a range of impacts on the quality of analyses that use the data. To provide a sound basis for research on this issue in databases of nucleotide sequences, we have developed new, large-scale validated collections of duplicates, which can be used to test the effectiveness of duplicate detection methods. Previous collections were either designed primarily to test efficiency, or contained only a limited number of duplicates of limited kinds. To date, duplicate detection methods have been evaluated on separate, inconsistent benchmarks, leading to results that cannot be compared and, due to limitations of the benchmarks, of questionable generality. In this study, we present three nucleotide sequence database benchmarks, based on information drawn from a range of resources, including information derived from mapping to two data sections within the UniProt Knowledgebase (UniProtKB), UniProtKB/Swiss-Prot and UniProtKB/TrEMBL. Each benchmark has distinct characteristics. We quantify these characteristics and argue for their complementary value in evaluation. The benchmarks collectively contain a vast number of validated biological duplicates; the largest has nearly half a billion duplicate pairs (although this is probably only a tiny fraction of the total that is present). They are also the first benchmarks targeting the primary nucleotide databases. The records include the 21 most heavily studied organisms in molecular biology research. Our quantitative analysis shows that duplicates in the different benchmarks, and in different organisms, have different characteristics. It is thus unreliable to evaluate duplicate detection methods against any single benchmark. For example, the benchmark derived from UniProtKB/Swiss-Prot mappings identifies more diverse types of duplicates, showing the importance of expert curation, but is limited to coding sequences. Overall, these benchmarks form a resource that we believe will be of great value for development and evaluation of the duplicate detection or record linkage methods that are required to help maintain these essential resources. DATABASE URL: : https://bitbucket.org/biodbqual/benchmarks.
-
ItemNo Preview AvailableOn the impact of initialisation strategies on Maximum Flow algorithm performance(PERGAMON-ELSEVIER SCIENCE LTD, 2024-03)
-
ItemNo Preview AvailableDirective Explanations for Actionable Explainability in Machine Learning Applications(ASSOC COMPUTING MACHINERY, 2023-12)In this article, we show that explanations of decisions made by machine learning systems can be improved by not only explaining why a decision was made but also explaining how an individual could obtain their desired outcome. We formally define the concept of directive explanations (those that offer specific actions an individual could take to achieve their desired outcome), introduce two forms of directive explanations (directive-specific and directive-generic), and describe how these can be generated computationally. We investigate people’s preference for and perception toward directive explanations through two online studies, one quantitative and the other qualitative, each covering two domains (the credit scoring domain and the employee satisfaction domain). We find a significant preference for both forms of directive explanations compared to non-directive counterfactual explanations. However, we also find that preferences are affected by many aspects, including individual preferences and social factors. We conclude that deciding what type of explanation to provide requires information about the recipients and other contextual information. This reinforces the need for a human-centered and context-specific approach to explainable AI.