Part Of Adaptation Research Articles

Partial Domain Adaptation by Progressive Sample Learning of Shared Classes

Partial Domain Adaptation by Progressive Sample Learning of Shared Classes

Partial domain adaption based prediction calibration methodology for fault detection and diagnosis of chillers under variable operational condition scenarios

Data-driven fault detection and diagnosis (FDD) for building energy system has drawn much attention from both industry and academy due to their ability to learn complex relationships from big data. Most existing FDD studies only focus on the model diagnosis accuracy where the operational conditions of test dataset are similar with training dataset. However, it is rare for online operational data to be similar with offline training data. To better simulate the online environment in the aspect of operational conditions, a novel online data simulation method, which is named as variable operational conditions test, is proposed in this study. Furthermore, in order to improve the poor diagnosis performance of FDD models under variable operational condition scenarios, an unsupervised partial domain adaption algorithm is employed to calibrate the predictions of the FDD model within a short period of time. A comprehensive data experiment was conduct based on ASHARE RP-1043. The experimental results indicate the prediction calibration method yields decent performance improvement compared with three traditional data-driven-based FDD models, the highest and overall improvements are 55.07% and 14.5%, respectively. The research outcomes provide practical guidelines for the online application of FDD models when the training data cannot cover the whole sample space.

Partial Transfer Ensemble Learning Framework: A Method for Intelligent Diagnosis of Rotating Machinery Based on an Incomplete Source Domain.

Most cross-domain intelligent diagnosis approaches presume that the health states in training datasets are consistent with those in testing. However, it is usually difficult and expensive to collect samples under all failure states during the training stage in actual engineering; this causes the training dataset to be incomplete. These existing methods may not be favorably implemented with an incomplete training dataset. To address this problem, a novel deep-learning-based model called partial transfer ensemble learning framework (PT-ELF) is proposed in this paper. The major procedures of this study consist of three steps. First, the missing health states in the training dataset are supplemented by another dataset. Second, since the training dataset is drawn from two different distributions, a partial transfer mechanism is explored to train a weak global classifier and two partial domain adaptation classifiers. Third, a particular ensemble strategy combines these classifiers with different classification ranges and capabilities to obtain the final diagnosis result. Two case studies are used to validate our method. Results indicate that our method can provide robust diagnosis results based on an incomplete source domain under variable working conditions.

Sustainable optimization of global aquatic omega-3 supply chain could substantially narrow the nutrient gap

Omega-3 EPA and DHA fatty acids are vital for human health, but current human nutritional requirements are greater than supply. This nutrient gap is poised to increase as demand increases and the abundance of aquatic foods and the amount of omega-3 they contain may dwindle due to climate change and overfishing. Identifying and mitigating loss and inefficiencies across the global aquatic supply chain has great potential for narrowing this nutrient gap. Here, using an optimization model, we show that omega-3 supply to humans could potentially increase by as much as 50% (reaching 630 kt y−1) compared to present baseline by shifting feed inputs to produce species that have the highest omega-3 content per feed input (i.e. carp and crustaceans), diverting other production flows towards direct wild fish consumption, improving byproduct utilization, and reducing waste at the retail and consumer level. We then discuss the implications of our findings by prioritizing policies and identifying demand- and supply-side interventions to realize these ambitious changes. This work emphasizes the urgency needed in managing aquatic resources towards greater utilization of resources and highlights the extent to which even partial adaptation of the measures we propose can have on narrowing the present and future nutrient gap as novel alternative sources of omega-3 become available on a larger scale.

GENTRIFICATION AS A PERSPECTIVE WAY TO RENEW INDUSTRIAL AND HISTORICAL AREAS OF THE CITY

The article provides a definition of the term "gentrification", explains its meaning and origin. It is determined that the founder of gentrification is the British social scientist Ruth Glass, who first described the process of gentrification in her book «London: Aspects of Change» in 1964. The paper analyzes and defines the main features and goals of gentrification. The concepts of "revitalization" and "gentrification" are compared and their common and individual features are highlighted. It is noted that the main goal of both processes, both gentrification and revitalization, is the renewal of the degrading historical former industrial areas of the city, with full or partial refunctionalization and adaptation to the modern needs of residents.
 The paper analyzes the current state and current problems of the industrial and historical territories of Kyiv. As a result, several main zones were identified, where old non-functioning industrial facilities are located. It has been found that over time, the "industrial model" of Kyiv city, formed in the industrial age, gradually turns into unprofitable, inefficient and neglected, which often leads to the degradation of city blocks and even entire districts. The author of the article proposes gentrification as a promising way to restore non-functioning industrial areas and abandoned industrial facilities.
 As an example of gentrification, author show the project of a new residential complex "Richert Park" in the Podolsky historical district of Kyiv. The article reveals the main aspects of the project and its main ideas. Based on the residential complex "Richert Park", an analysis was made of the revitalization of a historical heritage of architecture (Richert's factory). According to the technique that has been determined, the historical heritage is harmoniously combined with modern new buildings.

Dual-Weight Consistency-Induced Partial Domain Adaptation Network for Intelligent Fault Diagnosis of Machinery

Domain adaptation-based methods have been broadly developed for cross-domain fault diagnosis of machinery, in which the target and source label spaces are identical. However, a more challenging and practical diagnostic scenario in which the source label space subsumes target one is remaining to be solved. To address this issue, an innovative two-stage dual-weight consistency-induced partial domain adaptation (DCPDA) network, consisting of two feature extractors, a classifier, a domain discriminator, a dual-weight consistency-induced reweighting (DCR) module, and a Wasserstein distance based-domain adaptation (WDA) module, is proposed. In the pre-trained stage, a well-trained model on the source domain is acquired. In the fine-tuned stage, the DCR, which conducts a dual-weight consistency-induced weighting strategy, obtains a bi-level compound weight from both class-level and sample-level weights. Embedded with the bi-level weight, the WDA is encouraged to transfer the source knowledge with mutual agreement from double-level weights across domains. Accordingly, the DCPDA is capable of alleviating negative transfer dramatically and mapping the source diagnostic knowledge into the target domain selectively. Extensive experiments indicate the validity and advantage of the proposed DCPDA.

Partial Domain Adaptation Method Based on Class-Weighted Alignment for Fault Diagnosis of Rotating Machinery

Domain adaptation (DA)-based methods for fault diagnosis (FD) of rotating machinery have achieved impressive results in recent years. Most methods hold the assumption that the source domain (SD) and target domain (TD) share the same label space, which is not always satisfied in actual situations. A more practical scenario called partial domain adaptation (PDA) needs to be given more attention, where the transferable knowledge learned from a larger SD is applied to a smaller but relevant TD. A PDA method called class-weighted alignment-based transfer network (CWATN) is proposed in this paper to adapt to this scenario. A novel weighting method is designed to adapt the data distributions of shared classes. Except for weighted class-level alignment, global-level feature adaptation is also considered to learn more general transferable knowledge. Moreover, a domain discrepancy learning block is plugged in the shared classifier as a residual block, which could enforce the network to learn and measure the discrepancy between SD and TD explicitly, thus improving the result of DA. Three case studies are implemented to verify the superiority of CWATN. Results demonstrate that the proposed method could obtain better diagnostic performance than the selected competitive methods in the PDA scenario.

Coupling Adversarial Learning with Selective Voting Strategy for Distribution Alignment in Partial Domain Adaptation

In contrast to a standard closed-set domain adaptation task, partial domain adaptation setup caters to a realistic scenario by relaxing the identical label set assumption. The fact of source label set subsuming the target label set, however, introduces few additional obstacles as training on private source category samples thwart relevant knowledge transfer and mislead the classification process. To mitigate these issues, we devise a mechanism for strategic selection of highly confident target samples essential for the estimation of class-importance weights. Furthermore, we capture class-discriminative and domain-invariant features by coupling the process of achieving compact and distinct class distributions with an adversarial objective. Experimental findings over numerous cross-domain classification tasks demonstrate the potential of the proposed technique to deliver superior and comparable accuracy over existing methods.

Automatic field measurement of groundwater discharges in a tunnel

The paper describes a complex project of flow rate measurement applied for groundwater discharge into a tunnel, for both individual seepage points and the collection channel totalizing the flow. Various standard methods are used according to a large range of flow rate magnitudes (tipping bucket, danaid, V-shaped weir) with partial improvements and technical adaptations. Data collection is provided by individual measurement units with ethernet interface sending data to a database server through the internet. Experience of 10 years with several upgrades of either individual measurement or the electronic collection and transmission system is summarized.

Upper respiratory tract symptom risk in elite field hockey players during a dry run for the Tokyo Olympics

ABSTRACT The primary aim of this study was to examine if biomarker and/or self-reported data could predict upper respiratory tract symptom (URTS) risk in elite field hockey players. The secondary aim was to investigate the effect of the additional stressor ‘repeated heat exposure’ on measures of thermoregulation and immunity. A prospective cohort repeated measures study design was used to collect URTS, household illness, self-reported wellness, biomarker and thermoregulatory data from elite male field hockey players (n = 19), during an 8-week training and competition period that simulated the preparatory and competition phases of the 2020 Tokyo Olympics. Heat response testing (HRT) was performed at the beginning of the study period, following heat acclimation (HA) and following an intensified competition period (ICP) played in hot and humid conditions (27–37°C and 53–80% relative humidity). Univariate frailty analysis demonstrated that illness in players' households (Hazard ratio (HR: 4.90; p < 0.001)) and self-reported stress (HR: 0.63; p = 0.043) predicted players' risk for URTS. Additionally, low baseline resting salivary secretory immunoglobulin A (SIgA) concentration predicted players' “potential” URTS risk (p = 0.021). The additional stressor “repeated heat exposure” was found to facilitate partial thermoregulatory adaptation without attenuating resting immune functions. In conclusion, lifestyle and behavioural factors (i.e. household illness and stress) influenced players risk for URTS more so than sport-related stressors. Furthermore, repeated heat exposure did not appear to compromise players resting immunity. To assess athletes’ risk for URTS, baseline screening of SIgA concentration and regular monitoring of self-reported lifestyle and behavioural data are recommended. Highlights Self-reported illness in players’ households and higher self-reported stress significantly predicted increased upper respiratory tract symptom risk. Low baseline salivary secretory immunoglobulin A concentration predicted players “potential” URTS risk. Repeated heat exposures facilitated partial thermoregulatory adaptation without altering resting immunity.

Hierarchical domain adaptation with local feature patterns

Domain adaptation is proposed to generalize learning machines and address performance degradation of models that are trained from one specific source domain but applied to novel target domains. Existing domain adaptation methods focus on transferring holistic features whose discriminability is generally tailored to be source-specific and inferiorly generic to be transferable. As a result, standard domain adaptation on holistic features usually damages feature structures, especially local feature statistics, and deteriorates the learned discriminability. To alleviate this issue, we propose to transfer primitive local feature patterns, whose discriminability are shown to be inherently more sharable, and perform hierarchical feature adaptation. Concretely, we first learn a cluster of domain-shared local feature patterns and partition the feature space into cells. Local features are adaptively aggregated inside each cell to obtain cell features, which are further integrated into holistic features. To achieve fine-grained adaptations, we simultaneously perform alignment on local features, cell features and holistic features, within which process the local and cell features are aligned independently inside each cell to maintain the learned local structures and prevent negative transfer. Experimenting on typical one-to-one unsupervised domain adaptation for both image classification and action recognition tasks, partial domain adaptation, and domain-agnostic adaptation, we show that the proposed method achieves more reliable feature transfer by consistently outperforming state-of-the-art models and the learned domain-invariant features generalize well to novel domains.

Prescripción potencialmente inapropiada en adultos mayores en México

ABSTRACTOBJECTIVE To identify and quantify potentially inappropriate prescribing (prescripción potencialmente inapropiada, PPI) and other drug prescribing problems in public health care services in a population-based study at the three existing levels of complexity in Mexico.METHODS Descriptive analysis of the Study on Satisfaction of Users of the Social Protection System in Health 2014–2016, prescription and drug supply section, to obtain the prevalence of PPI in older adults (≥ 65 years), based on Beers, STOPP, Prescrire and BSP listings using AM (older adults) prescription indicators, one for each listing.RESULTS Most older adults (67%) were prescribed at least one medication, with a mean of 2.7 medications per prescription. The PPI prevalence was 74% according to the BSP criteria, 67% according to the STOPP listing, 59% with the Beer criteria, and 20% with Prescrire. The most frequent PPI prescriptions were NSAIDs, vasodilators and sulfonylureas.CONCLUSIONS The use of PPIs in AM is high in Mexico. The higher prevalence found in this study may reflect the use of a source with population representativeness. The partial use and adaptations of the criteria make difficult comparing the studies; however, the STOPP criteria are the ones with the highest prevalence, as they cover a greater number of drugs and their use is more common in the first level of care.

Robotic-Assisted Partial Nephrectomy: Techniques to Improve Clinical Outcomes.

To summarize current options available for robot-assisted partial nephrectomy RECENT FINDINGS: Partial nephrectomy (PN) is a standard treatment option for management of cT1 renal masses. It may be carried out by multiple approaches. Robot-assisted (RA) PN is one such option. The goal of treatment is both correct oncological (negative surgical margins) and functional (preservation of sufficient amount of renal parenchyma of the operated kidney) outcome. Appropriate outcomes depend on multiple factors. There are many, but among others tumor characteristics (size, location, i.e., tumor complexity), patient baseline renal function, patient comorbidities, and performance status etc. Based on all these, the surgeon adapts the intervention for each mass/patient by preoperative planning, absence/use/duration of warm or cold ischemia, perioperative imaging, resection technique adapted to tumor location and depth of invasion, use of hemostatics, type and degree of renal parenchymal closure and others details. Nephroprotective agents have not shown efficacy so far. It should not be forgotten that surgeon's experience plays a key role in the achievement of good results. Although multiple factors have a role in the RA partial nephrectomy, surgeon experience and adaptation of technique of intervention have the crucial role in the achievement of both functional and oncological results.

Choroideremia Gene Therapy.

Choroideremia Gene Therapy.

Multi-source transfer learning network to complement knowledge for intelligent diagnosis of machines with unseen faults

Most of the current successes of deep transfer learning-based fault diagnosis require two assumptions: 1) the health state set of source machines should overlap that of target machines; 2) the number of target machine samples is balanced across health states. However, such assumptions are unrealistic in engineering scenarios, where target machines suffer from fault types that are not seen in source machines and the target machines are mostly in a healthy state with only occasional faults. As a result, the diagnostic knowledge from source machines may not cover all fault types of target machines nor address imbalanced target samples. Therefore, we propose a framework, called a multi-source transfer learning network (MSTLN), to aggregate and transfer diagnostic knowledge from multiple source machines by combining multiple partial distribution adaptation sub-networks (PDA-Subnets) and a multi-source diagnostic knowledge fusion module. The former weights target samples by counter-balancing factors to jointly adapt partial distributions of source and target pairs, and the latter releases negative effects due to discrepancy among multiple source machines and further fuses diagnostic decisions output from multiple PDA-Subnets. Two case studies demonstrate that MSTLN can reduce the misdiagnosis rate and obtain better transfer performance for imbalanced target samples than other conventional methods.

Multi-scene ancient Chinese text recognition with deep coupled alignments

The task of multi-scene ancient Chinese text recognition (MACR) is challenging due to large-scale categories, high intra-class variance and inter-class similarity and complicated backgrounds. Little effort has been devoted to MACR research due to insufficient datasets and language barrier. Because the sub-dataset generation process of sub-dataset is mutually blind, there are discrepancies in the class category number, deep feature representation and class center distribution after the dataset statistics and character analysis are performed. The general deep learning method that assumes that data are independent and identically distributed is inappropriate. The deep coupled alignments (CA) module based on domain adaptation is presented to alleviate domain and class center shifts. In addition, a cross-domain fusion (CF) module is proposed to mitigate negative transfer in partial domain adaptation by updating the target domain with the full-class and augmenting the source domain with pseudo labeled samples. Extensive experiments of the proposed method are conducted, and the results illustrate the superiority of CA–CF to previous methods in terms of the model size and recognition accuracy.

Learning Target-Domain-Specific Classifier for Partial Domain Adaptation.

Unsupervised domain adaptation (UDA) aims at reducing the distribution discrepancy when transferring knowledge from a labeled source domain to an unlabeled target domain. Previous UDA methods assume that the source and target domains share an identical label space, which is unrealistic in practice since the label information of the target domain is agnostic. This article focuses on a more realistic UDA scenario, i.e., partial domain adaptation (PDA), where the target label space is subsumed to the source label space. In the PDA scenario, the source outliers that are absent in the target domain may be wrongly matched to the target domain (technically named negative transfer), leading to performance degradation of UDA methods. This article proposes a novel target-domain-specific classifier learning-based domain adaptation (TSCDA) method. TSCDA presents a soft-weighed maximum mean discrepancy criterion to partially align feature distributions and alleviate negative transfer. Also, it learns a target-specific classifier for the target domain with pseudolabels and multiple auxiliary classifiers to further address the classifier shift. A module named peers-assisted learning is used to minimize the prediction difference between multiple target-specific classifiers, which makes the classifiers more discriminant for the target domain. Extensive experiments conducted on three PDA benchmark data sets show that TSCDA outperforms other state-of-the-art methods with a large margin, e.g., 4% and 5.6% averagely on Office-31 and Office-Home, respectively.

Virus Host Jumping Can Be Boosted by Adaptation to a Bridge Plant Species.

Understanding biological mechanisms that regulate emergence of viral diseases, in particular those events engaging cross-species pathogens spillover, is becoming increasingly important in virology. Species barrier jumping has been extensively studied in animal viruses, and the critical role of a suitable intermediate host in animal viruses-generated human pandemics is highly topical. However, studies on host jumping involving plant viruses have been focused on shifting intra-species, leaving aside the putative role of “bridge hosts” in facilitating interspecies crossing. Here, we take advantage of several VPg mutants, derived from a chimeric construct of the potyvirus Plum pox virus (PPV), analyzing its differential behaviour in three herbaceous species. Our results showed that two VPg mutations in a Nicotiana clevelandii-adapted virus, emerged during adaptation to the bridge-host Arabidopsis thaliana, drastically prompted partial adaptation to Chenopodium foetidum. Although both changes are expected to facilitate productive interactions with eIF(iso)4E, polymorphims detected in PPV VPg and the three eIF(iso)4E studied, extrapolated to a recent VPg:eIF4E structural model, suggested that two adaptation ways can be operating. Remarkably, we found that VPg mutations driving host-range expansion in two non-related species, not only are not associated with cost trade-off constraints in the original host, but also improve fitness on it.

Evolutionary mismatch along salinity gradients in a Neotropical water strider.

The evolution of local adaptation is crucial for the in situ persistence of populations in changing environments. However, selection along broad environmental gradients could render local adaptation difficult, and might even result in maladaptation. We address this issue by quantifying fitness trade‐offs (via common garden experiments) along a salinity gradient in two populations of the Neotropical water strider Telmatometra withei—a species found in both fresh (FW) and brackish (BW) water environments across Panama. We found evidence for local adaptation in the FW population in its home FW environment. However, the BW population showed only partial adaptation to the BW environment, with a high magnitude of maladaptation along naturally occurring salinity gradients. Indeed, its overall fitness was ~60% lower than that of the ancestral FW population in its home environment, highlighting the role of phenotypic plasticity, rather than local adaptation, in high salinity environments. This suggests that populations seemingly persisting in high salinity environments might in fact be maladapted, following drastic changes in salinity. Thus, variable selection imposed by salinization could result in evolutionary mismatch, where the fitness of a population is displaced from its optimal environment. Understanding the fitness consequences of persisting in fluctuating salinity environments is crucial to predict the persistence of populations facing increasing salinization. It will also help develop evolutionarily informed management strategies in the context of global change.

Manejo extrahospitalario de la fibrilación auricular y su Adecuación a las guías de práctica clínica

Objetivo: Estudiar la variabilidad del manejo de la fibrilación auricular (FA) por equipos de emergencias y su adherencia a las guías de tratamiento. Material y método. Estudio descriptivo de episodios de FA atendidos en dos años en Cádiz. Recogemos 236 pacientes. Analizamos antecedentes, actuación del equipo y resultados de control de frecuencia y ritmo. Construimos un índice sintético de adecuación terapéutica (ISAT) con rango de -3 a 3. Resultados. 236 pacientes, 68% varones, las mujeres más jóvenes (66 vs 75 años); 43.8% fueron primeros episodios. El fármaco más usado es la amiodarona. El 16,1% de las asistencias presentaron adherencia completa (ISAT 3) y el 63,1%, adecuación parcial positiva (ISAT 1-2); la mayor variabilidad clínica se produjo en la bradicardización. La adecuación fue: 62,3% para anticoagulación, 68.2% para bradicardización y 56.8% para cardioversión. Existió un 40,7% de control de frecuencia y un 96% de derivación hospitalaria. Conclusión. Debemos aumentar el porcentaje de anticoagulación y empleo de bloqueadores-beta y calcioantagonistas en sustitución de amiodarona. El ISAT permite mejorar la adecuación a las GPC que facilita controlar síntomas, frecuencia y reducir estancias en los servicios de urgencia.