J. Carter, K. Dennerl, K. Kuntz, W. Dunn, D. Bodewits, C. Jackman, S. Sembay, G. Branduardi-Raymont, T. Deskins, D. Koutroumpa, R. Kraft, C. Lisse, S. McEntee, S. Wolk, F. S. Porter
Mars provides our local analogue for unmagnetized terrestrial planets and is thus key to understanding the habitability of exoplanets. The lack of a global magnetic field means that the atmosphere interacts directly with the solar wind, causing significant loss of the atmosphere. While in situ measurements provide a wealth of detailed local information, they are limited in deriving the global picture. In contrast, remote X-ray observations can provide important global instantaneous coverage over multiple seasons and sampling different solar wind. Previous XMM-Newton observations have detected significant flux via the solar wind charge exchange emission (SWCX) mechanism from an extended planetary halo, and from atmospheric fluorescence. In contrast, Chandra observations only detected a low-luminosity disk and a faint halo. It is postulated that these observational differences are due to transient solar wind with increased heavy ion fractions. Here, we present simulated spectra for the proposed NASA mission Line Emission Mapper, of both halo and disk regions, under quiet and transient solar wind. We show that even under moderate solar wind conditions, both SWCX and fluorescence emission lines are readily detected above the background, providing new insights into the loss of planetary atmospheres and the molecular composition of less well-characterised atmospheric abundances.
火星为我们提供了当地未磁化陆地行星的类似物，因此是了解系外行星可居住性的关键。全球磁场的缺乏意味着大气层会直接与太阳风相互作用，导致大气层的大量损失。虽然现场测量提供了大量详细的局部信息，但在推导全球情况方面却很有限。相比之下，远程 X 射线观测可以在多个季节提供重要的全球瞬时覆盖范围，并对不同的太阳风进行采样。先前的 XMM-牛顿观测通过太阳风电荷交换发射（SWCX）机制从扩展的行星晕和大气荧光中探测到了大量通量。相比之下，钱德拉观测仅探测到一个低亮度的圆盘和一个微弱的光晕。据推测，这些观测差异是由于重离子分数增加的瞬态太阳风造成的。在此，我们展示了拟议中的 NASA 任务 "线发射成像仪 "在静态和瞬态太阳风条件下对光晕和盘区的模拟光谱。我们的研究表明，即使在中等太阳风条件下，SWCX 和荧光发射线也很容易在背景之上被探测到，从而为行星大气的损耗和特征不太明显的大气丰度的分子组成提供了新的见解。
{"title":"The exosphere of Mars can be tracked by a high-spectral resolution telescope, such as the Line Emission Mapper","authors":"J. Carter, K. Dennerl, K. Kuntz, W. Dunn, D. Bodewits, C. Jackman, S. Sembay, G. Branduardi-Raymont, T. Deskins, D. Koutroumpa, R. Kraft, C. Lisse, S. McEntee, S. Wolk, F. S. Porter","doi":"10.1093/rasti/rzae033","DOIUrl":"https://doi.org/10.1093/rasti/rzae033","url":null,"abstract":"\u0000 Mars provides our local analogue for unmagnetized terrestrial planets and is thus key to understanding the habitability of exoplanets. The lack of a global magnetic field means that the atmosphere interacts directly with the solar wind, causing significant loss of the atmosphere. While in situ measurements provide a wealth of detailed local information, they are limited in deriving the global picture. In contrast, remote X-ray observations can provide important global instantaneous coverage over multiple seasons and sampling different solar wind. Previous XMM-Newton observations have detected significant flux via the solar wind charge exchange emission (SWCX) mechanism from an extended planetary halo, and from atmospheric fluorescence. In contrast, Chandra observations only detected a low-luminosity disk and a faint halo. It is postulated that these observational differences are due to transient solar wind with increased heavy ion fractions. Here, we present simulated spectra for the proposed NASA mission Line Emission Mapper, of both halo and disk regions, under quiet and transient solar wind. We show that even under moderate solar wind conditions, both SWCX and fluorescence emission lines are readily detected above the background, providing new insights into the loss of planetary atmospheres and the molecular composition of less well-characterised atmospheric abundances.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"5 4","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141921167","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
J. Weston, K. W. Smith, S. Smartt, J. Tonry, H. Stevance
We present a Convolutional Neural Network (CNN) for use in the Real-Bogus classification of transient detections made by the Asteroid Terrestrial Impact Last Alert System (ATLAS) and subsequent efforts to improve performance since initial development. In transient detection surveys the number of alerts made outstrips the capacity for human scanning, necessitating the use of machine learning aids to reduce the number of false positives presented to annotators. We take a sample of recently annotated data from each of the three operating ATLAS telescope with ∼340,000 real (known transients) and ∼1,030,000 bogus detections per model. We retrained the CNN architecture with these data specific to each ATLAS unit, achieving a median False Positive Rate (FPR) of 0.72 per cent for a 1.00 per cent missed detection rate. Further investigations indicate if we reduce the input image size it results in increases to the false positive rate. Finally architecture adjustments and comparisons to contemporary CNNs indicate our retrained classifier is providing an optimal FPR. We conclude that the periodic retraining and readjustment of classification models on survey data can yield significant improvements as data drift arising from changes to optical and detector performance can lead to new features in the model and subsequent deteriorations in performance.
{"title":"Training a convolutional neural network for real-bogus classification in the ATLAS survey","authors":"J. Weston, K. W. Smith, S. Smartt, J. Tonry, H. Stevance","doi":"10.1093/rasti/rzae027","DOIUrl":"https://doi.org/10.1093/rasti/rzae027","url":null,"abstract":"\u0000 We present a Convolutional Neural Network (CNN) for use in the Real-Bogus classification of transient detections made by the Asteroid Terrestrial Impact Last Alert System (ATLAS) and subsequent efforts to improve performance since initial development. In transient detection surveys the number of alerts made outstrips the capacity for human scanning, necessitating the use of machine learning aids to reduce the number of false positives presented to annotators. We take a sample of recently annotated data from each of the three operating ATLAS telescope with ∼340,000 real (known transients) and ∼1,030,000 bogus detections per model. We retrained the CNN architecture with these data specific to each ATLAS unit, achieving a median False Positive Rate (FPR) of 0.72 per cent for a 1.00 per cent missed detection rate. Further investigations indicate if we reduce the input image size it results in increases to the false positive rate. Finally architecture adjustments and comparisons to contemporary CNNs indicate our retrained classifier is providing an optimal FPR. We conclude that the periodic retraining and readjustment of classification models on survey data can yield significant improvements as data drift arising from changes to optical and detector performance can lead to new features in the model and subsequent deteriorations in performance.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"12 9","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141654278","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The presence of a planetary companion around its host star has been repeatedly linked with stellar properties, affecting the likelihood of sub-stellar object formation and stability in the protoplanetary disc, thus presenting a key challenge in exoplanet science. Furthermore, abundance and stellar parameter datasets tend to be incomplete, which limits the ability to infer distributional characteristics harnessing the entire dataset. This work aims to develop a methodology using machine learning and multi-objective optimisation for reliable imputation for subsequent comparison tests and host star recommendation. It integrates fuzzy clustering for imputation and ML classification of hosts and comparison stars into an evolutionary multi-objective optimisation algorithm. We test several candidates for the classification model, starting with a binary classification for giant planet hosts. Upon confirmation that the XGBoost algorithm provides the best performance, we interpret the performance of both the imputation and classification modules for binary classification. The model is extended to handle multi-label classification for low-mass planets and planet multiplicity. Constraints on the model’s use and feature/sample selection are given, outlining strengths and limitations. We conclude that the careful use of this technique for host star recommendation will be an asset to future missions and the compilation of necessary target lists.
宿主恒星周围的行星伴星的存在一再与恒星特性联系在一起，影响着亚恒星天体形成的可能性和原行星盘的稳定性，从而给系外行星科学带来了关键的挑战。此外，丰度和恒星参数数据集往往不完整，这限制了利用整个数据集推断分布特征的能力。这项工作旨在利用机器学习和多目标优化开发一种方法，为后续对比测试和宿主星推荐提供可靠的估算。它将用于估算的模糊聚类以及主机和对比恒星的 ML 分类集成到进化多目标优化算法中。我们测试了几个候选分类模型，首先是巨行星宿主的二元分类。在确认 XGBoost 算法提供了最佳性能之后，我们对双星分类的估算和分类模块的性能进行了解释。我们对模型进行了扩展，以处理低质量行星和行星多度的多标签分类。我们给出了模型使用和特征/样本选择的约束条件，概述了其优势和局限性。我们的结论是，谨慎使用这一技术来推荐主星将是未来任务和编制必要目标列表的宝贵财富。
{"title":"Exoplanet host star classification: Multi-Objective Optimisation of incomplete stellar abundance data","authors":"Miguel A Zammit, Josef Borg, Kristian Zarb Adami","doi":"10.1093/rasti/rzae020","DOIUrl":"https://doi.org/10.1093/rasti/rzae020","url":null,"abstract":"\u0000 The presence of a planetary companion around its host star has been repeatedly linked with stellar properties, affecting the likelihood of sub-stellar object formation and stability in the protoplanetary disc, thus presenting a key challenge in exoplanet science. Furthermore, abundance and stellar parameter datasets tend to be incomplete, which limits the ability to infer distributional characteristics harnessing the entire dataset. This work aims to develop a methodology using machine learning and multi-objective optimisation for reliable imputation for subsequent comparison tests and host star recommendation. It integrates fuzzy clustering for imputation and ML classification of hosts and comparison stars into an evolutionary multi-objective optimisation algorithm. We test several candidates for the classification model, starting with a binary classification for giant planet hosts. Upon confirmation that the XGBoost algorithm provides the best performance, we interpret the performance of both the imputation and classification modules for binary classification. The model is extended to handle multi-label classification for low-mass planets and planet multiplicity. Constraints on the model’s use and feature/sample selection are given, outlining strengths and limitations. We conclude that the careful use of this technique for host star recommendation will be an asset to future missions and the compilation of necessary target lists.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"28 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141358680","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Feargus A J Abernethy, Hannah Chinnery, Robert Lindner, Simeon J Barber
Ever increasing interest in the Moon, not only for scientific but also commercial and prospecting purposes, requires a more streamlined and reproduceable approach to issues such as the sealing of sample handling ovens, in contrast to the mission-specific mechanisms which have tended to prevail in the past. A test breadboard has been designed and built in order to evaluate the leak rates of different oven sealing concepts and materials within the context of the ProSPA instrument being developed for the European Space Agency. Sealing surface geometries based on a simple 90° knife-edge, and two widely used vacuum fitting standards (VCR® and ConFlat®) have been tested using PTFE gaskets in vacuum across a temperature range of -100°C to 320°C, equivalent to a projected -100°C to 1000°C sample heating range in the ProSPA ovens. The impact of using glass- and carbon- filled PTFE has also been investigated, as has the effect of dust coverage of JSC-1A lunar simulant up to 9 per cent by area. The best combination of properties appears to be unfilled PTFE, compressed between two 90° knife-edges with a confining force of ∼ 400 N. This can produce a leak rates within the 10−7 Pa.m3.s−1 range or better regardless of the level of dust applied within the experimental constraints. A strong temperature-dependence on the leak rate is identified, meaning that careful oven design will be required to minimise the temperature at the seal interface even within the operational temperature range PTFE itself.
{"title":"PTFE as a viable sealing material for lightweight mass spectrometry ovens in dusty extraterrestrial environments","authors":"Feargus A J Abernethy, Hannah Chinnery, Robert Lindner, Simeon J Barber","doi":"10.1093/rasti/rzae003","DOIUrl":"https://doi.org/10.1093/rasti/rzae003","url":null,"abstract":"\u0000 Ever increasing interest in the Moon, not only for scientific but also commercial and prospecting purposes, requires a more streamlined and reproduceable approach to issues such as the sealing of sample handling ovens, in contrast to the mission-specific mechanisms which have tended to prevail in the past. A test breadboard has been designed and built in order to evaluate the leak rates of different oven sealing concepts and materials within the context of the ProSPA instrument being developed for the European Space Agency. Sealing surface geometries based on a simple 90° knife-edge, and two widely used vacuum fitting standards (VCR® and ConFlat®) have been tested using PTFE gaskets in vacuum across a temperature range of -100°C to 320°C, equivalent to a projected -100°C to 1000°C sample heating range in the ProSPA ovens. The impact of using glass- and carbon- filled PTFE has also been investigated, as has the effect of dust coverage of JSC-1A lunar simulant up to 9 per cent by area. The best combination of properties appears to be unfilled PTFE, compressed between two 90° knife-edges with a confining force of ∼ 400 N. This can produce a leak rates within the 10−7 Pa.m3.s−1 range or better regardless of the level of dust applied within the experimental constraints. A strong temperature-dependence on the leak rate is identified, meaning that careful oven design will be required to minimise the temperature at the seal interface even within the operational temperature range PTFE itself.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"4 3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139846257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Feargus A J Abernethy, Hannah Chinnery, Robert Lindner, Simeon J Barber
Ever increasing interest in the Moon, not only for scientific but also commercial and prospecting purposes, requires a more streamlined and reproduceable approach to issues such as the sealing of sample handling ovens, in contrast to the mission-specific mechanisms which have tended to prevail in the past. A test breadboard has been designed and built in order to evaluate the leak rates of different oven sealing concepts and materials within the context of the ProSPA instrument being developed for the European Space Agency. Sealing surface geometries based on a simple 90° knife-edge, and two widely used vacuum fitting standards (VCR® and ConFlat®) have been tested using PTFE gaskets in vacuum across a temperature range of -100°C to 320°C, equivalent to a projected -100°C to 1000°C sample heating range in the ProSPA ovens. The impact of using glass- and carbon- filled PTFE has also been investigated, as has the effect of dust coverage of JSC-1A lunar simulant up to 9 per cent by area. The best combination of properties appears to be unfilled PTFE, compressed between two 90° knife-edges with a confining force of ∼ 400 N. This can produce a leak rates within the 10−7 Pa.m3.s−1 range or better regardless of the level of dust applied within the experimental constraints. A strong temperature-dependence on the leak rate is identified, meaning that careful oven design will be required to minimise the temperature at the seal interface even within the operational temperature range PTFE itself.
{"title":"PTFE as a viable sealing material for lightweight mass spectrometry ovens in dusty extraterrestrial environments","authors":"Feargus A J Abernethy, Hannah Chinnery, Robert Lindner, Simeon J Barber","doi":"10.1093/rasti/rzae003","DOIUrl":"https://doi.org/10.1093/rasti/rzae003","url":null,"abstract":"\u0000 Ever increasing interest in the Moon, not only for scientific but also commercial and prospecting purposes, requires a more streamlined and reproduceable approach to issues such as the sealing of sample handling ovens, in contrast to the mission-specific mechanisms which have tended to prevail in the past. A test breadboard has been designed and built in order to evaluate the leak rates of different oven sealing concepts and materials within the context of the ProSPA instrument being developed for the European Space Agency. Sealing surface geometries based on a simple 90° knife-edge, and two widely used vacuum fitting standards (VCR® and ConFlat®) have been tested using PTFE gaskets in vacuum across a temperature range of -100°C to 320°C, equivalent to a projected -100°C to 1000°C sample heating range in the ProSPA ovens. The impact of using glass- and carbon- filled PTFE has also been investigated, as has the effect of dust coverage of JSC-1A lunar simulant up to 9 per cent by area. The best combination of properties appears to be unfilled PTFE, compressed between two 90° knife-edges with a confining force of ∼ 400 N. This can produce a leak rates within the 10−7 Pa.m3.s−1 range or better regardless of the level of dust applied within the experimental constraints. A strong temperature-dependence on the leak rate is identified, meaning that careful oven design will be required to minimise the temperature at the seal interface even within the operational temperature range PTFE itself.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"119 22","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139786455","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We examine critically recent claims for the presence of above–atmosphere optical transients in publicly–available digitised scans of Schmidt telescope photographic plate material derived from the National Geographic Society–Palomar Observatory Sky Survey. We employ the publicly available SuperCOSMOS Sky Survey catalogues to examine statistically the morphology of the sources. We develop a simple, objective and automated image classification scheme based on a random forest decision tree classifier. We find that the putative transients are likely to be spurious artefacts of the photographic emulsion. We suggest a possible cause of the appearance of these images as resulting from the copying procedure employed to disseminate glass copy survey atlas sets in the era before large–scale digitisation programmes.
{"title":"On the nature of apparent transient sources on the National Geographic Society–Palomar Observatory Sky Survey glass copy plates","authors":"Nigel Hambly, Adam Blair","doi":"10.1093/rasti/rzae004","DOIUrl":"https://doi.org/10.1093/rasti/rzae004","url":null,"abstract":"\u0000 We examine critically recent claims for the presence of above–atmosphere optical transients in publicly–available digitised scans of Schmidt telescope photographic plate material derived from the National Geographic Society–Palomar Observatory Sky Survey. We employ the publicly available SuperCOSMOS Sky Survey catalogues to examine statistically the morphology of the sources. We develop a simple, objective and automated image classification scheme based on a random forest decision tree classifier. We find that the putative transients are likely to be spurious artefacts of the photographic emulsion. We suggest a possible cause of the appearance of these images as resulting from the copying procedure employed to disseminate glass copy survey atlas sets in the era before large–scale digitisation programmes.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"19 11","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139822396","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We examine critically recent claims for the presence of above–atmosphere optical transients in publicly–available digitised scans of Schmidt telescope photographic plate material derived from the National Geographic Society–Palomar Observatory Sky Survey. We employ the publicly available SuperCOSMOS Sky Survey catalogues to examine statistically the morphology of the sources. We develop a simple, objective and automated image classification scheme based on a random forest decision tree classifier. We find that the putative transients are likely to be spurious artefacts of the photographic emulsion. We suggest a possible cause of the appearance of these images as resulting from the copying procedure employed to disseminate glass copy survey atlas sets in the era before large–scale digitisation programmes.
{"title":"On the nature of apparent transient sources on the National Geographic Society–Palomar Observatory Sky Survey glass copy plates","authors":"Nigel Hambly, Adam Blair","doi":"10.1093/rasti/rzae004","DOIUrl":"https://doi.org/10.1093/rasti/rzae004","url":null,"abstract":"\u0000 We examine critically recent claims for the presence of above–atmosphere optical transients in publicly–available digitised scans of Schmidt telescope photographic plate material derived from the National Geographic Society–Palomar Observatory Sky Survey. We employ the publicly available SuperCOSMOS Sky Survey catalogues to examine statistically the morphology of the sources. We develop a simple, objective and automated image classification scheme based on a random forest decision tree classifier. We find that the putative transients are likely to be spurious artefacts of the photographic emulsion. We suggest a possible cause of the appearance of these images as resulting from the copying procedure employed to disseminate glass copy survey atlas sets in the era before large–scale digitisation programmes.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"10 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139882188","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract Probabilistic classification of unassociated Fermi-LAT sources using machine learning methods has an implicit assumption that the distributions of associated and unassociated sources are the same as a function of source parameters, which is not the case for the Fermi-LAT catalogs. The problem of different distributions of training and testing (or target) datasets as a function of input features (covariates) is known as the covariate shift. In this paper, we, for the first time, quantitatively estimate the effect of the covariate shift on the multi-class classification of Fermi-LAT sources. We introduce sample weights proportional to the ratio of unassociated to associated source probability density functions so that associated sources in areas, which are densely populated with unassociated sources, have more weight than the sources in areas with few unassociated sources. We find that the covariate shift has relatively little effect on the predicted probabilities, i.e. the training can be performed either with weighted or with unweighted samples, which is generally expected for the covariate shift problems. The main effect of the covariate shift is on the estimated performance of the classification. Depending on the class, the covariate shift can lead up to 10 – 20% reduction in precision and recall compared to the estimates, where the covariate shift is not taken into account.
{"title":"Effect of covariate shift on multi-class classification of Fermi-LAT sources","authors":"Dmitry V Malyshev","doi":"10.1093/rasti/rzad053","DOIUrl":"https://doi.org/10.1093/rasti/rzad053","url":null,"abstract":"Abstract Probabilistic classification of unassociated Fermi-LAT sources using machine learning methods has an implicit assumption that the distributions of associated and unassociated sources are the same as a function of source parameters, which is not the case for the Fermi-LAT catalogs. The problem of different distributions of training and testing (or target) datasets as a function of input features (covariates) is known as the covariate shift. In this paper, we, for the first time, quantitatively estimate the effect of the covariate shift on the multi-class classification of Fermi-LAT sources. We introduce sample weights proportional to the ratio of unassociated to associated source probability density functions so that associated sources in areas, which are densely populated with unassociated sources, have more weight than the sources in areas with few unassociated sources. We find that the covariate shift has relatively little effect on the predicted probabilities, i.e. the training can be performed either with weighted or with unweighted samples, which is generally expected for the covariate shift problems. The main effect of the covariate shift is on the estimated performance of the classification. Depending on the class, the covariate shift can lead up to 10 – 20% reduction in precision and recall compared to the estimates, where the covariate shift is not taken into account.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"132 37","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136351685","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mohammad Hassan Hassanshahi, Marcin Jastrzebski, Sarah Malik, Ofer Lahav
Abstract Galaxy morphology, a key tracer of the evolution of a galaxy’s physical structure, has motivated extensive research on machine learning techniques for efficient and accurate galaxy classification. The emergence of quantum computers has generated optimism about the potential for significantly improving the accuracy of such classifications by leveraging the large dimensionality of quantum Hilbert space. This paper presents a quantum-enhanced support vector machine algorithm for classifying galaxies based on their morphology. The algorithm requires the computation of a kernel matrix, a task that is performed on a simulated quantum computer using a quantum circuit conjectured to be intractable on classical computers. The result shows similar performance between classical and quantum-enhanced support vector machine algorithms. For a training size of 40k, the receiver operating characteristic curve for differentiating ellipticals and spirals has an under-curve area (ROC AUC) of 0.946 ± 0.005 for both classical and quantum-enhanced algorithms. Additionally, we demonstrate for a small dataset that the performance of a noise-mitigated quantum SVM algorithm on a quantum device is in agreement with simulation. Finally, a necessary condition for achieving a potential quantum advantage is presented. This investigation is among the very first applications of quantum machine learning in astronomy and highlights their potential for further application in this field.
{"title":"A quantum-enhanced support vector machine for galaxy classification","authors":"Mohammad Hassan Hassanshahi, Marcin Jastrzebski, Sarah Malik, Ofer Lahav","doi":"10.1093/rasti/rzad052","DOIUrl":"https://doi.org/10.1093/rasti/rzad052","url":null,"abstract":"Abstract Galaxy morphology, a key tracer of the evolution of a galaxy’s physical structure, has motivated extensive research on machine learning techniques for efficient and accurate galaxy classification. The emergence of quantum computers has generated optimism about the potential for significantly improving the accuracy of such classifications by leveraging the large dimensionality of quantum Hilbert space. This paper presents a quantum-enhanced support vector machine algorithm for classifying galaxies based on their morphology. The algorithm requires the computation of a kernel matrix, a task that is performed on a simulated quantum computer using a quantum circuit conjectured to be intractable on classical computers. The result shows similar performance between classical and quantum-enhanced support vector machine algorithms. For a training size of 40k, the receiver operating characteristic curve for differentiating ellipticals and spirals has an under-curve area (ROC AUC) of 0.946 ± 0.005 for both classical and quantum-enhanced algorithms. Additionally, we demonstrate for a small dataset that the performance of a noise-mitigated quantum SVM algorithm on a quantum device is in agreement with simulation. Finally, a necessary condition for achieving a potential quantum advantage is presented. This investigation is among the very first applications of quantum machine learning in astronomy and highlights their potential for further application in this field.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"31 8","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135430180","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A Spurio Mancini, M M Docherty, M A Price, J D McEwen
Abstract Comparison of appropriate models to describe observational data is a fundamental task of science. The Bayesian model evidence, or marginal likelihood, is a computationally challenging, yet crucial, quantity to estimate to perform Bayesian model comparison. We introduce a methodology to compute the Bayesian model evidence in simulation-based inference (SBI) scenarios (also often called likelihood-free inference). In particular, we leverage the recently proposed learned harmonic mean estimator and exploit the fact that it is decoupled from the method used to generate posterior samples, i.e. it requires posterior samples only, which may be generated by any approach. This flexibility, which is lacking in many alternative methods for computing the model evidence, allows us to develop SBI model comparison techniques for the three main neural density estimation approaches, including neural posterior estimation (NPE), neural likelihood estimation (NLE), and neural ratio estimation (NRE). We demonstrate and validate our SBI evidence calculation techniques on a range of inference problems, including a gravitational wave example. Moreover, we further validate the accuracy of the learned harmonic mean estimator, implemented in the harmonic software, in likelihood-based settings. These results highlight the potential of harmonic as a sampler-agnostic method to estimate the model evidence in both likelihood-based and simulation-based scenarios.
{"title":"Bayesian model comparison for simulation-based inference","authors":"A Spurio Mancini, M M Docherty, M A Price, J D McEwen","doi":"10.1093/rasti/rzad051","DOIUrl":"https://doi.org/10.1093/rasti/rzad051","url":null,"abstract":"Abstract Comparison of appropriate models to describe observational data is a fundamental task of science. The Bayesian model evidence, or marginal likelihood, is a computationally challenging, yet crucial, quantity to estimate to perform Bayesian model comparison. We introduce a methodology to compute the Bayesian model evidence in simulation-based inference (SBI) scenarios (also often called likelihood-free inference). In particular, we leverage the recently proposed learned harmonic mean estimator and exploit the fact that it is decoupled from the method used to generate posterior samples, i.e. it requires posterior samples only, which may be generated by any approach. This flexibility, which is lacking in many alternative methods for computing the model evidence, allows us to develop SBI model comparison techniques for the three main neural density estimation approaches, including neural posterior estimation (NPE), neural likelihood estimation (NLE), and neural ratio estimation (NRE). We demonstrate and validate our SBI evidence calculation techniques on a range of inference problems, including a gravitational wave example. Moreover, we further validate the accuracy of the learned harmonic mean estimator, implemented in the harmonic software, in likelihood-based settings. These results highlight the potential of harmonic as a sampler-agnostic method to estimate the model evidence in both likelihood-based and simulation-based scenarios.","PeriodicalId":500957,"journal":{"name":"RAS Techniques and Instruments","volume":"18 s1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135430324","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：