A simulated annealing-based maximum-margin clustering algorithm
- Seifollahi, Sattar, Bagirov, Adil, Borzeshi, Ehsan, Piccardi, Massimo
- Authors: Seifollahi, Sattar , Bagirov, Adil , Borzeshi, Ehsan , Piccardi, Massimo
- Date: 2019
- Type: Text , Journal article
- Relation: Computational Intelligence Vol. 35, no. 1 (2019), p. 23-41
- Full Text:
- Reviewed:
- Description: Maximum-margin clustering is an extension of the support vector machine (SVM) to clustering. It partitions a set of unlabeled data into multiple groups by finding hyperplanes with the largest margins. Although existing algorithms have shown promising results, there is no guarantee of convergence of these algorithms to global solutions due to the nonconvexity of the optimization problem. In this paper, we propose a simulated annealing-based algorithm that is able to mitigate the issue of local minima in the maximum-margin clustering problem. The novelty of our algorithm is twofold, ie, (i) it comprises a comprehensive cluster modification scheme based on simulated annealing, and (ii) it introduces a new approach based on the combination of k-means++ and SVM at each step of the annealing process. More precisely, k-means++ is initially applied to extract subsets of the data points. Then, an unsupervised SVM is applied to improve the clustering results. Experimental results on various benchmark data sets (of up to over a million points) give evidence that the proposed algorithm is more effective at solving the clustering problem than a number of popular clustering algorithms.
- Authors: Seifollahi, Sattar , Bagirov, Adil , Borzeshi, Ehsan , Piccardi, Massimo
- Date: 2019
- Type: Text , Journal article
- Relation: Computational Intelligence Vol. 35, no. 1 (2019), p. 23-41
- Full Text:
- Reviewed:
- Description: Maximum-margin clustering is an extension of the support vector machine (SVM) to clustering. It partitions a set of unlabeled data into multiple groups by finding hyperplanes with the largest margins. Although existing algorithms have shown promising results, there is no guarantee of convergence of these algorithms to global solutions due to the nonconvexity of the optimization problem. In this paper, we propose a simulated annealing-based algorithm that is able to mitigate the issue of local minima in the maximum-margin clustering problem. The novelty of our algorithm is twofold, ie, (i) it comprises a comprehensive cluster modification scheme based on simulated annealing, and (ii) it introduces a new approach based on the combination of k-means++ and SVM at each step of the annealing process. More precisely, k-means++ is initially applied to extract subsets of the data points. Then, an unsupervised SVM is applied to improve the clustering results. Experimental results on various benchmark data sets (of up to over a million points) give evidence that the proposed algorithm is more effective at solving the clustering problem than a number of popular clustering algorithms.
Fermentation and the microbial community of Japanese koji and miso : a review
- Allwood, Joanne, Wakeling, Lara, Bean, David
- Authors: Allwood, Joanne , Wakeling, Lara , Bean, David
- Date: 2021
- Type: Text , Journal article , Review
- Relation: Journal of Food Science Vol. 86, no. 6 (2021), p. 2194-2207
- Full Text:
- Reviewed:
- Description: Abstract: Miso is a well-known traditional Japanese fermented food, with a characteristic savory flavor and aroma, known predominately as the seasoning in miso soup. Miso production involves a two-stage fermentation, where first a mold, such as Aspergillus oryzae, is inoculated onto a substrate to make koji. A subsequent fermentation, this time by bacteria and yeast, occurs when the koji is added to a salt and soybean mash, with the miso left to ferment for up to 2 years. The microbial community of miso is considered essential to the development of the unique taste, texture, and nutritional profile of miso. Despite the importance of microorganisms in the production of miso, very little research has been undertaken to characterize and describe the microbial process. In this review, we provide an overview of the two-stage fermentation process, describe what is currently known about the microbial communities involved and consider any potential health benefits associated with the consumption of miso, along with food safety concerns. As the popularity of miso continues to expand globally and is produced under new environmental conditions, understanding the microbiological processes involved will assist to ensure that global production of miso is safe as well as delicious. © 2021 Institute of Food Technologists®
- Authors: Allwood, Joanne , Wakeling, Lara , Bean, David
- Date: 2021
- Type: Text , Journal article , Review
- Relation: Journal of Food Science Vol. 86, no. 6 (2021), p. 2194-2207
- Full Text:
- Reviewed:
- Description: Abstract: Miso is a well-known traditional Japanese fermented food, with a characteristic savory flavor and aroma, known predominately as the seasoning in miso soup. Miso production involves a two-stage fermentation, where first a mold, such as Aspergillus oryzae, is inoculated onto a substrate to make koji. A subsequent fermentation, this time by bacteria and yeast, occurs when the koji is added to a salt and soybean mash, with the miso left to ferment for up to 2 years. The microbial community of miso is considered essential to the development of the unique taste, texture, and nutritional profile of miso. Despite the importance of microorganisms in the production of miso, very little research has been undertaken to characterize and describe the microbial process. In this review, we provide an overview of the two-stage fermentation process, describe what is currently known about the microbial communities involved and consider any potential health benefits associated with the consumption of miso, along with food safety concerns. As the popularity of miso continues to expand globally and is produced under new environmental conditions, understanding the microbiological processes involved will assist to ensure that global production of miso is safe as well as delicious. © 2021 Institute of Food Technologists®
Outfoxing the fox : effect of prey odor on fox behavior in a pastoral landscape
- Andrewartha, Tim, Evans, Maldwyn, Batson, William, Manning, Adrian, Barton, Philip
- Authors: Andrewartha, Tim , Evans, Maldwyn , Batson, William , Manning, Adrian , Barton, Philip
- Date: 2021
- Type: Text , Journal article
- Relation: Conservation Science and Practice Vol. 3, no. 12 (2021), p.
- Full Text:
- Reviewed:
- Description: Invasive mammalian predators have had a devastating effect on native species globally. The European red fox (Vulpes vulpes) is one such species where it has been introduced in Australia. A novel but unexplored tactic to reduce the impact of mammalian predators is the use of unrewarded prey odors to undermine the effectiveness of olfactory hunting behavior. To test the viability of unrewarded prey odors in an applied setting we investigated how foxes responded to the odors of three different prey species. We used the odors of two locally extinct native Australian marsupials; the eastern quoll (a smaller carnivore) and eastern bettong (a fungivore), and the European rabbit, an introduced herbivore. Conducting our research over a period of 3 weeks in a pastoral environment in South-eastern Australia, we used video observations of foxes' behaviors, as they encountered the different odors. We found a reduction in the number of fox visits to bettong odors in the third week. In contrast, we observed a sustained number of visits to rabbit odors. Foxes also spent more time investigating rabbit odors and displayed longer durations of vigilance behavior at quoll odors. Our results support the hypothesis that the exposure of wild foxes to unrewarded odors of novel prey species can reduce their interest in these odors, which might translate to a reduction in predation pressure. Our results also suggest, however, that olfactory pre-exposure may not be as effective at reducing fox interest in a competitor species' odor. © 2021 The Authors. Conservation Science and Practice published by Wiley Periodicals LLC. on behalf of Society for Conservation Biology
- Authors: Andrewartha, Tim , Evans, Maldwyn , Batson, William , Manning, Adrian , Barton, Philip
- Date: 2021
- Type: Text , Journal article
- Relation: Conservation Science and Practice Vol. 3, no. 12 (2021), p.
- Full Text:
- Reviewed:
- Description: Invasive mammalian predators have had a devastating effect on native species globally. The European red fox (Vulpes vulpes) is one such species where it has been introduced in Australia. A novel but unexplored tactic to reduce the impact of mammalian predators is the use of unrewarded prey odors to undermine the effectiveness of olfactory hunting behavior. To test the viability of unrewarded prey odors in an applied setting we investigated how foxes responded to the odors of three different prey species. We used the odors of two locally extinct native Australian marsupials; the eastern quoll (a smaller carnivore) and eastern bettong (a fungivore), and the European rabbit, an introduced herbivore. Conducting our research over a period of 3 weeks in a pastoral environment in South-eastern Australia, we used video observations of foxes' behaviors, as they encountered the different odors. We found a reduction in the number of fox visits to bettong odors in the third week. In contrast, we observed a sustained number of visits to rabbit odors. Foxes also spent more time investigating rabbit odors and displayed longer durations of vigilance behavior at quoll odors. Our results support the hypothesis that the exposure of wild foxes to unrewarded odors of novel prey species can reduce their interest in these odors, which might translate to a reduction in predation pressure. Our results also suggest, however, that olfactory pre-exposure may not be as effective at reducing fox interest in a competitor species' odor. © 2021 The Authors. Conservation Science and Practice published by Wiley Periodicals LLC. on behalf of Society for Conservation Biology
Soil reconstruction after mining fails to restore soil function in an Australian arid woodland
- Duncan, Corrine, Good, Megan, Sluiter, Ian, Cook, Simon, Schultz, Nick
- Authors: Duncan, Corrine , Good, Megan , Sluiter, Ian , Cook, Simon , Schultz, Nick
- Date: 2020
- Type: Text , Journal article
- Relation: Restoration Ecology Vol. 28, no. S1 (2020), p. A35-A43
- Full Text:
- Reviewed:
- Description: The biogeochemical properties of soils drive ecosystem function and vegetation dynamics, and hence soil restoration after mining should aim to reinstate the soil properties and hydrological dynamics of remnant ecosystems. The aim of this study is to assess soil structure in two vegetation types in an arid ecosystem, and to understand how these soil properties compare to a reconstructed soil profile after mining. In an arid ecosystem in southeast Australia, soil samples were collected at five depths (to 105 cm) from remnant woodland and shrubland sites, and sites either disturbed or totally reconstructed after mining. We assessed soil physico-chemical properties and microbial activity. Soils in the remnant arid ecosystem had coarse-textured topsoils that overlay clay horizons, which allows water to infiltrate and avoid evaporation, but also slows drainage to deeper horizons. Conversely, reconstructed soils had high sand content at subsoil horizons and high bulk density and compaction at surface layers (0–20 cm). Reconstructed soils had topsoils with higher pH and electrical conductivity. The reconstructed soils did not show increased microbial activity with time since restoration. Overall, the reconstructed soil horizons were not organized in a way that allowed rainfall infiltration and water storage, as is imperative to arid-zone ecosystem function. Future restoration efforts in arid ecosystems should focus on increasing sand content of soils near the surface, to reduce evaporative water loss and improve soil quality and plant health. © 2020 Society for Ecological Restoration
- Authors: Duncan, Corrine , Good, Megan , Sluiter, Ian , Cook, Simon , Schultz, Nick
- Date: 2020
- Type: Text , Journal article
- Relation: Restoration Ecology Vol. 28, no. S1 (2020), p. A35-A43
- Full Text:
- Reviewed:
- Description: The biogeochemical properties of soils drive ecosystem function and vegetation dynamics, and hence soil restoration after mining should aim to reinstate the soil properties and hydrological dynamics of remnant ecosystems. The aim of this study is to assess soil structure in two vegetation types in an arid ecosystem, and to understand how these soil properties compare to a reconstructed soil profile after mining. In an arid ecosystem in southeast Australia, soil samples were collected at five depths (to 105 cm) from remnant woodland and shrubland sites, and sites either disturbed or totally reconstructed after mining. We assessed soil physico-chemical properties and microbial activity. Soils in the remnant arid ecosystem had coarse-textured topsoils that overlay clay horizons, which allows water to infiltrate and avoid evaporation, but also slows drainage to deeper horizons. Conversely, reconstructed soils had high sand content at subsoil horizons and high bulk density and compaction at surface layers (0–20 cm). Reconstructed soils had topsoils with higher pH and electrical conductivity. The reconstructed soils did not show increased microbial activity with time since restoration. Overall, the reconstructed soil horizons were not organized in a way that allowed rainfall infiltration and water storage, as is imperative to arid-zone ecosystem function. Future restoration efforts in arid ecosystems should focus on increasing sand content of soils near the surface, to reduce evaporative water loss and improve soil quality and plant health. © 2020 Society for Ecological Restoration
- «
- ‹
- 1
- ›
- »