Long-term Information Research Articles

Weakly supervised video anomaly detection based on hyperbolic space

In recent years, there has been a proliferation of weakly supervised methods in the field of video anomaly detection. Despite significant progress in existing research, these efforts have primarily focused on addressing this issue within Euclidean space. Conducting weakly supervised video anomaly detection in Euclidean space imposes a fundamental limitation by constraining the ability to model complex patterns due to the dimensionality constraints of the embedding space and lacking the capacity to model long-term contextual information. This inadequacy can lead to misjudgments of anomalous events due to insufficient video representation. However, hyperbolic space has shown significant potential for modeling complex data, offering new insights. In this paper, we rethink weakly supervised video anomaly detection with a novel perspective: transforming video features from Euclidean space into hyperbolic space may enable the network to learn implicit relationships in normal and anomalous videos, thereby enhancing its ability to effectively distinguish between them. Finally, to validate our approach, we conducted extensive experiments on the UCF-Crime and XD-Violence datasets. Experimental results show that our method not only has the lowest number of parameters but also achieves state-of-the-art performance on the XD-Violence dataset using only RGB information.

Long-Term Implicit Epigenetic Stress Information in the Enteric Nervous System and its Contribution to Developing and Perpetuating IBS.

Psychiatric and mood disorders may play an important role in the development and persistence of irritable bowel syndrome (IBS). Previously, we hypothesized that stress-induced implicit memories may persist throughout life via epigenetic processes in the enteric nervous system (ENS), independent of the central nervous system (CNS). These epigenetic memories in the ENS may contribute to developing and perpetuating IBS. Here, we further elaborate on our earlier hypothesis. That is, during pregnancy, maternal prenatal stresses perturb the HPA axis and increase circulating cortisol levels, which can affect the maternal gut microbiota. Maternal cortisol can cross the placental barrier and increase cortisol-circulating levels in the fetus. This leads to dysregulation of the HPA axis, affecting the gut microbiota, microbial metabolites, and intestinal permeability in the fetus. Microbial metabolites, such as short-chain fatty acids (which also regulate the development of fetal ENS), can modulate a range of diseases by inducing epigenetic changes. These mentioned processes suggest that stress-related, implicit, long-term epigenetic memories may be programmed into the fetal ENS during pregnancy. Subsequently, this implicit epigenetic stress information from the fetal ENS could be conveyed to the CNS through the bidirectional microbiota-gut-brain axis (MGBA), leading to perturbed functional connectivity among various brain networks and the dysregulation of affective and pain processes.

Percutaneous ultrasound-guided anchor wire placement aids in the intraoperative localization of nonpalpable, superficial foreign bodies and abscesses in dogs.

To describe and evaluate the use of preoperative percutaneous ultrasound-guided anchor wire placement to aid intraoperative localization of superficial foreign bodies and abscesses in dogs. 11 dogs. In a retrospective observational study, the medical records of dogs that underwent surgical exploration of superficial abscesses, guided by anchor wire, between 2018 and 2023 were reviewed for clinical and histopathological findings and complications. Owners or veterinarians were contacted to collect long-term follow-up information. 11 dogs were included. Superficial swelling was the most common clinical presentation. Computed tomography and ultrasound revealed an abscess cavity and suspected foreign body in 9 dogs and an abscess cavity without evidence of a foreign body in 2 dogs. Anchor wires were placed in close proximity to the foreign body or inside the abscess. All documented foreign bodies were successfully located and retrieved. Two dogs suffered minor postoperative complications. No major intra- or postoperative complications were documented. One dog displayed recurrence of clinical signs, but no further surgical management was required. Preoperative percutaneous placement of an anchor wire via ultrasound guidance was successful in aiding intraoperative localization of nonpalpable abscesses and retrieval of foreign bodies. This technique may decrease surgical time, minimize the surgical approach required, and increase the likelihood of successful localization.

Prognostic utility of heart-type fatty acid binding protein in patients with cardiac amyloidosis

Abstract Background Heart-type fatty acid binding protein (H-FABP) is a biomarker that has been shown to provide long-term prognostic information in patients with heart failure (HF), yet its role in patients with cardiac amyloidosis (CA) is unknown. Purpose We aim to investigate the association of H-FABP levels with prognosis in patients with CA. Methods Consecutive patients diagnosed with cardiac amyloidosis from October 2015 to May 2022 at Heart Failure Center, of our hospital were enrolled. Patients were categorized into 2 categories based on H-FABP levels measured in the baseline blood sample: < or = 7 ng/ml (low H-FABP), >7 ng/ml (high H-FABP). Univariate and multivariate Cox models were used to identify predictors of survival. Kaplan-Meier analysis was performed to compare survival between patients with low or high H-FABP. Results Overall, ninety-five patients were included. H-FABP was elevated (>7 ng/ml) in 37 patients (39%). Patients with high H-FABP were more likely to present with higher levels of high-sensitivity troponin-T (hs-TNT)(0.17ng/mL vs. 0.11ng/mL, p =0.01) and higher levels of N-terminal pro-B-type natriuretic peptide (NT-proBNP) (18098.76pg/mL vs. 6211.69pg/mL, p<0.001). Patients were followed for an average follow-up period of 440 days, in which 53 (56%) patients died. Univariate and multivariate analysis demonstrated that H-FABP levels (HR 1.48, 95% CI 1.03-2.13, per 1-SD greater) was independently associated with prognosis, even after adjusting for NT-proBNP, hs-TNT and renal function. Conclusions Elevation of H-FABP is associated with an increased risk of death in patients with cardiac amyloidosis. H-FABP can serve as a novel biomarker independent of other established predictors for prognosis in cardiac amyloidosis.Kaplan-Meier curves for overall survival

Characteristics of applicants with a first-time recognition of need for long-term care

Abstract Background The increasing number of care-dependent individuals requires approaches to prevent care dependency or reduce loss of independence. Long-term care assessments can provide valuable insights into this. In Germany, the Medical Service is responsible for the long-term care assessment (care level) that focuses on limitations in independent living. Objectives The aim of this contribution is to describe initial applicants with identified need for long-term care. Materials and methods The nationwide database consists of long-term care assessments conducted by the German Medical Service of health insured people aged 60 and above who received a care level (1=smallest till 5= largest limitations) for the first time in 2021. Long-term care-relevant information is analyzed descriptively. Results 339,486 individuals with an average age of 79.6 years (SD: 8.4) were analyzed. The age distribution shows that the 80-89 age group comprises half of the sample (49.9%), whereas those aged 90+ encompass the smallest proportion of the sample (9.3%). More women than men received a care level (59.0% women vs. 41.0% men), with women predominant in care level 1-2 and men in the higher care level 3-5. Overall, approximately half of the sample received care level 2, 32.4 % received care level 1. Care levels 3-5 were assessed less frequently (16.2 % vs. 4.8 % vs. 1.7 %). The most frequent care-relevant diagnoses were senility (R54), polyarthrosis (M15), and dementia (F03). Conclusions Over three quarters of initial applicants are classified into the lowest care levels (1 and 2). Accordingly, tertiary prevention measures are recommended, as these may delay the further progression into a higher care level. Key messages • Over three quarters of initial applicants for long-term care are classified into the lowest care levels. Accordingly, tertiary prevention measures may recommended to delay further care progression. • The most frequent care-relevant diagnoses were senility (R54), polyarthrosis (M15), and dementia (F03).

Model-Informed Drug Development-Based Approval of Intravenous Secukinumab for the Treatment of Adult Patients with Active Psoriatic Arthritis, Active Ankylosing Spondylitis, and Active Non-Radiographic Axial Spondyloarthritis.

On October 6, 2023, the US Food and Drug Administration (FDA) approved an intravenous (IV) formulation and dosage of Cosentyx® (secukinumab), for the treatment of adult patients with active psoriatic arthritis (PsA), active ankylosing spondylitis (AS), and active non-radiographic axial spondyloarthritis (nr-axSpA) with objective signs of inflammation. Clinical pharmacokinetics (PK), efficacy, and short-term placebo-controlled safety data were available from clinical studies (NCT04156620 and NCT04209205) with the to-be-marketed IV formulation using a maintenance dosage 3 mg/kg every 4 weeks (q4w), which was different from the dose approved (1.75 mg/kg q4w). The IV dosage of 3 mg/kg utilized in these two trials resulted in exposures (Cmax,ss) significantly higher than those for the approved subcutaneous (SC) regimens. Further, there is limited long-term safety information available for this 3 mg/kg q4w IV dose. To address this important limitation, a model-informed drug development (MIDD) approach was employed to leverage available clinical PK, efficacy, and safety data from the secukinumab development program to identify a maintenance IV dosing regimen, 1.75 mg/kg IV q4w, that better approximated the relevant SC secukinumab exposures for which efficacy and safety have been established. The MIDD analyses were used to support approval of this IV dosing regimen not directly studied in the indications sought for licensure, PsA, AS, and nr-AxSpA.

Speaker Recognition System Using Hybrid of MFCC and RCNN with HCO Algorithm Optimization

Though there are advancements in speaker recognition technology, available systems often fail to correctly recognize speakers especially in noisy environments. The use of Mel-frequency cepstral coefficients (MFCC) has been improved using Convolutional Neural Networks (CNN) yet difficulties in achieving high accuracies still exists. Hybrid algorithms combining MFCC and Region-based Convolutional Neural Networks (RCNN) have been found to be promising. In this research features from speech signals were extracted for speaker recognition, to denoise the signals, design and develop a DFT-based denoising system using spectrum subtraction and to develop a speaker recognition method for the Verbatim Transcription using MFCC. The DFT was used to transform the sampled audio signal waveform into a frequency-domain signal. RCNN was used to model the characteristics of speakers based on their voice samples, and to classify them into different categories or identities. The novelty of the research was that it used MFCC integrated with RCNN and optimized with Host-Cuckoo Optimization (HCO) algorithm. HCO algorithm is capable of further weight optimization through the process of generating fit cuckoos for best weights. It also captured the temporal dependencies and long-term information. The system was tested and validated on audio recordings from different personalities from the National Assembly of Kenya. The results were compared with the actual identity of the speakers to confirm accuracy. The performance of the proposed approach was compared with two other existing speaker recognition the traditional approaches being MFCC-CNN and Linear Predictive Coefficients (LPC)-CNN. The comparison was based the Equal Error Rate (EER), False Rejection Rate (FRR), False Match Rate (FMR), and True Match Rate (TMR). Results show that the proposed algorithm outperformed the others in maintaining a lowest EER, FMR, FRR and highest TMR.

Colonic torsion and volvulus in dogs is associated with a low mortality rate and good long-term outcome.

To describe clinical findings, complications, and short- and long-term outcomes associated with colonic torsion and volvulus in dogs. 28 client-owned dogs. Multi-institutional, retrospective study. Medical records were searched for dogs of any age, body weight, sex, and breed that underwent surgery for colonic torsion and volvulus. Collected data included signalment, previous history, preoperative findings, time until surgery, intraoperative findings, postoperative complications, length of hospitalization, survival to discharge, and outcomes. 28 dogs were included. Thirteen of 28 dogs (46.4%) had preexisting gastrointestinal conditions. Nine of 28 dogs (32.1%) had a gastropexy performed prior to presentation. Ten dogs (35.7%) were found to have a resolution of colonic torsion and volvulus at the time of the surgery. All but 1 dog (27 of 28 [96.4%]) survived to discharge. Two dogs died during the postoperative period, yielding a mortality rate of 7.1%. Postoperative complications were noted in 9 dogs (9 of 28 [32.1%]). Long-term follow-up information was available in 16 of 28 dogs (57%). Among 16 dogs with at least 6 months' follow-up, all dogs (16 of 16 [100%]) were alive at 6 months postoperatively. Two dogs developed mesenteric torsion after the initial surgery. Dogs with colonic torsion and volvulus undergoing surgery can have an excellent survival-to-discharge ratio with a low mortality rate. Surgeons should not be prompted to euthanize or assume a guarded prognosis solely on the basis of the intraoperative appearance of the bowel and should consider all factors prior to making decisions. Owners should be informed of the risk of developing further torsional diseases after surgery.

Application of a novel visual representation method for urban noise source monitoring

Increasing urban mobility has led to a rise in the number of vehicles, resulting in urban noise pollution, which can cause health problems. Conventional methods for noise control can be adopted to reduce noise near the sources. However, this is not cost-effective to reduce overall noise level in the urban area. Therefore, it is preferable to encourage drivers to voluntarily reduce noise by applying methods such as cracking down on road noise sources. To this end, a real-time urban noise source monitoring system is needed, but processing large amounts of data can be difficult. In this study, a novel visual representation method, DoAgram, is applied for urban noise source monitoring. DoAgram can represent the time, frequency, and spatial information of the sound source within a 2D color graph. Therefore, the long-term spatial information of the sound source acquired from the sensor array node can be transmitted as a single image. Also, one can restore spatial information of the source from an image through a decoding process corresponding to the encoding method. One can find that the proposed method can be utilized as a database that can efficiently conduct urban noise source monitoring.

Safety of Medications Used to Treat Autoimmune Rheumatic Diseases During Pregnancy and Lactation.

Autoimmune rheumatic diseases (ARDs) often affect women during their reproductive years, and early studies of pregnancy in these patients reported high rates of adverse outcomes. Continuation or initiation of safe and effective medications in the preconception period is beneficial for maintaining or achieving disease quiescence throughout pregnancy thereby improving both maternal and pregnancy outcomes. The European Alliance of Associations for Rheumatology, the American College of Rheumatology, and the British Society for Rheumatology have published recommendations and guidelines regarding management of ARDs during pregnancy. The American College of Obstetricians and Gynecologists and the American Gastroenterological Association have also provided guidance statements with relevant recommendations. This review provides an overview of available recommendations for medication use in ARD pregnancy, with discussion of safety considerations for maternal and fetal well-being. Medications considered compatible with pregnancy include hydroxychloroquine, sulfasalazine, azathioprine, cyclosporine, tacrolimus, and TNF inhibitors. Methotrexate, mycophenolate, leflunomide, and cyclophosphamide should be avoided before and during pregnancy. Other medications, most of them newer, are largely discouraged for use in pregnancy due to inadequate data or concerns for neonatal immunosuppression, including non-TNF biologics and small molecule therapies. Further investigation is needed regarding effects of non-TNF biologics, biosimilars, and small molecules in pregnancy. Important efforts for the future will include improved methodologies to gather critical safety data, with consideration of inclusion of pregnant women in clinical trials, a complex and controversial issue. Long-term information on outcomes in offspring of treated women is lacking for many of these medications.

A hybrid deep learning-based short-term forecast model for ionospheric foF2 in East Asia region

The short-term forecast of the critical frequency of the ionospheric F2 layer (foF2) is particularly challenging due to its nonlinear and non-stationary characteristics. To improve the short-term forecast accuracy of foF2, we propose a short-term hybrid deep learning model that integrates the improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) and long short-term memory (LSTM) algorithms. The proposed model leverages the ICEEMDAN decomposition algorithm to expand foF2 time series data into multi-dimensional space and utilizes the LSTM algorithm to preserve long-term data information, capturing detailed changes in foF2 parameters and obtaining internal information about foF2. The 2014 and 2017 foF2 data from four observation stations located in Mohe, Beijing, Jeju, and Kokubunji in the mid-latitudes of East Asia were utilized for modeling and 1-hour forecast analysis. The forecast values of the proposed model align closely with the measured data, exhibiting minimal fluctuation in forecast error and effectively tracking the changing trend of foF2. The average root mean square error (RMSE) across the four stations in 2014 is 0.40 MHz, with a relative RMSE (RRMSE) of 7.11 % and a coefficient of determination (R2) of 0.98. The average RMSE in 2017 is 0.43 MHz, RRMSE is 7.68 %, and R2 is 0.92. The proposed model reveals significant improvements in the forecast accuracy of the comparison model compared with the CCIR, URSI, NN, and LSTM models. Especially compared with the CCIR and URSI models in the International Reference Ionosphere (IRI), demonstrating superior real-time tracking capability for foF2 and effectively addressing the challenge of poor short-term forecast accuracy observed in the IRI model.

RFNet: Multivariate long sequence time-series forecasting based on recurrent representation and feature enhancement

Multivariate time series exhibit complex patterns and structures involving interactions among multiple variables and long-term temporal dependencies, making multivariate long sequence time series forecasting (MLSTF) exceptionally challenging. Despite significant progress in Transformer-based methods in the MLSTF domain, many models still rely on stacked encoder–decoder architectures to capture complex time series patterns. This leads to increased computational complexity and overlooks spatial pattern information in multivariate time series, thereby limiting the model’s performance. To address these challenges, we propose RFNet, a lightweight model based on recurrent representation and feature enhancement. We partition the time series into fixed-size subsequences to retain local contextual temporal pattern information and cross-variable spatial pattern information. The recurrent representation module employs gate attention mechanisms and memory units to capture local information of the subsequences and obtain long-term correlation information of the input sequence by integrating information from different memory units. Meanwhile, we utilize a shared multi-layer perceptron (MLP) to capture global pattern information of the input sequence. The feature enhancement module explicitly extracts complex spatial patterns in the time series by transforming the input sequence. We validate the performance of RFNet on ten real-world datasets. The results demonstrate an improvement of approximately 55.3% over state-of-the-art MLSTF models, highlighting its significant advantage in addressing multivariate long sequence time series forecasting problems.

Long-term communication of aftershock forecasts: the Canterbury Earthquake Sequence in Aotearoa, New Zealand

On 14 February 2016, a magnitude (M)5.7 earthquake struck in Christchurch, Aotearoa, New Zealand. The shaking caused damage to historic facades, power outages, cliff collapses, rock falls, liquefaction but no reported injuries or fatalities. This Valentine’s Day earthquake was an aftershock in the Canterbury earthquake sequence (CES), which began on 4 September 2010 with the M7.1 Darfield Earthquake and included the destructive and fatal M6.2 Christchurch aftershock on 22 February 2011. This study eight months after the Valentine’s Day earthquake and six years after the initiation of the CES is the first to explore long-term aftershock forecast information and communication needs. The exploratory study also aimed to gather feedback on aftershock scenarios, an alternate form of communication of the forecast.The qualitative study involved workshops with emergency managers, public health officials, and members of the public in Christchurch. Key findings for long-term communication throughout an earthquake sequence include: 1. divergent earthquake experiences affect aftershock communication response and information needs; 2. understanding aftershock sequence behavior is foundational to sense-making when large aftershocks occur; 3. strategic earthquake sequence updates from the trusted science agency and local agencies could serve as important reminders for earthquake preparedness; 4. communication of aftershock forecast uncertainty could aid with both the credibility of the information and living with uncertainty, and 5. inclusion of impact information and preparedness advice into aftershock forecast scenarios could provide links to actionable information. The paper derives implications for research and practice of long-term communications during an aftershock sequence.

LS-Net: lightweight segmentation network for dermatological epidermal segmentation in optical coherence tomography imaging.

Optical coherence tomography (OCT) can be an important tool for non-invasive dermatological evaluation, providing useful data on epidermal integrity for diagnosing skin diseases. Despite its benefits, OCT's utility is limited by the challenges of accurate, fast epidermal segmentation due to the skin morphological diversity. To address this, we introduce a lightweight segmentation network (LS-Net), a novel deep learning model that combines the robust local feature extraction abilities of Convolution Neural Network and the long-term information processing capabilities of Vision Transformer. LS-Net has a depth-wise convolutional transformer for enhanced spatial contextualization and a squeeze-and-excitation block for feature recalibration, ensuring precise segmentation while maintaining computational efficiency. Our network outperforms existing methods, demonstrating high segmentation accuracy (mean Dice: 0.9624 and mean IoU: 0.9468) with significantly reduced computational demands (floating point operations: 1.131 G). We further validate LS-Net on our acquired dataset, showing its effectiveness in various skin sites (e.g., face, palm) under realistic clinical conditions. This model promises to enhance the diagnostic capabilities of OCT, making it a valuable tool for dermatological practice.

LLMDiff: Diffusion Model Using Frozen LLM Transformers for Precipitation Nowcasting.

Precipitation nowcasting, which involves the short-term, high-resolution prediction of rainfall, plays a crucial role in various real-world applications. In recent years, researchers have increasingly utilized deep learning-based methods in precipitation nowcasting. The exponential growth of spatiotemporal observation data has heightened interest in recent advancements such as denoising diffusion models, which offer appealing prospects due to their inherent probabilistic nature that aligns well with the complexities of weather forecasting. Successful application of diffusion models in rainfall prediction tasks requires relevant conditions and effective utilization to direct the forecasting process of the diffusion model. In this paper, we propose a probabilistic spatiotemporal model for precipitation nowcasting, named LLMDiff. The architecture of LLMDiff includes two networks: a conditional encoder-decoder network and a denoising network. The conditional network provides conditional information to guide the denoising network for high-quality predictions related to real-world earth systems. Additionally, we utilize a frozen transformer block from pre-trained large language models (LLMs) in the denoising network as a universal visual encoder layer, which enables the accurate estimation of motion trend by considering long-term temporal context information and capturing temporal dependencies within the frame sequence. Our experimental results demonstrate that LLMDiff outperforms state-of-the-art models on the SEVIR dataset.

Detection and characterization of aeolian sand strips on the beach using permanent laser scanning

Aeolian sand strips are bedforms that often form on the beach during high wind events. The sand strips can move across the beach resulting in sediment transport, but these sediment fluxes have not yet been quantified. Permanent laser scanning (PLS) provides an opportunity to gather long-term and detailed information about sand strips. Here, we present a method that detects sand strips in PLS reflectance images and extracts their wavelength, orientation, and height. The detection method is used on a month-long hourly dataset of Noordwijk beach, the Netherlands. The average sand strip wavelength (13 m), height (3 cm), and orientation (onshore-oblique) found in this dataset are consistent with previous research. Migration rates are obtained by examining successive images of sand strip patterns. Using the concurrent sand strip height, the sediment transport rate associated with sand strips is estimated. During a single 8-hour event, a total sediment flux of 0.16 m3/m is observed of which 0.07 m3/m in onshore direction. Furthermore, the detection method is shown to be applicable to other beaches by calibrating its thresholds. This opens a new possibility for future research to use PLS to investigate the behavior and sediment transport rates of aeolian sand strips.

Physical Function Trajectory among High-Functioning Long-Term Care Facility Residents: Utilizing Japanese National Data.

Physical function trajectory (PFT) is associated with mortality and hospitalization risks. We aimed to identify and compare the PFTs of newly admitted high-functioning older adults during their first six months at long-term care (LTC) facilities. In this multicenter retrospective cohort study, we included newly admitted high-functioning older adults (Barthel index > 60) from 47 Japanese LTC facilities. The primary outcome was physical function changes after admission. Data were collected from the Long-Term Care Information System for Evidence (LIFE), which monitored LTC facility residents' function between 1 January 2021 and 31 January 2022. A group-based trajectory model and binomial logistic regression analyses were applied to identify and compare residents' PFTs. Among the 718 residents included, the average age was 85.69 years and 64.5% were female. PFTs were classified as maintenance (66.0%), improvement (9.5%), slight decline (16.6%), and large decline (7.9%). The improvement group had significantly fewer residents who expressed a lack of interest in daily activities (odds ratio (OR) 0.45; 95% confidence interval (CI) 0.21-0.97) compared to the maintenance group. The large decline group had significantly more residents with a low BMI at admission (OR 2.42; 95% CI 1.29-4.55) and residents who did not use dentures (OR 0.49; 95% CI 0.26-0.95), compared to the maintenance group. Considering future PFTs may aid the development of care plans and the provision of appropriate interventions. Moreover, utilizing existing data has the potential to maintain residents' physical independence and enhance the quality of care without burdening residents themselves or staff.

Analysing the pedigree to identify undesirable losses of genetic diversity and to prioritize management decisions in captive breeding: a case study.

When prevention of species extinction is the priority, captive breeding is a key component in conservation programmes, allowing the recording of pedigree information in studbooks. The genealogical information registered in Cuvier's gazelle studbook between 1975 and 2023 was analysed to (a) assess if the implemented mating policy was successful in preserving the genetic background of the founders (1 male:3 females) in the present population, and b) improve future management and breeding decisions. Although the maternal contribution of one founder female was lost and the mean inbreeding of the total live population was high (0.305 ± 0.095), the breeding policy applied produced better results than expected from a population starting from four founders. It was successful in keeping the individual increase in inbreeding low (0.047 ± 0.021), and, notably, the inbreeding tended to decrease during the last three decades of the breeding programme, ensuring the viability of this highly inbred population. Historical dissemination of individuals among the zoos of Europe and North America caused population structuring and genetic differentiation of the live North American population. However, it did not risk the viability of the captive population. The average relatedness coefficients allowed the identification of individuals with underrepresented genotypes, which is relevant to plan future mating guidelines to keep the founders' representation balanced in the next generations. This study highlights the importance of keeping long-term pedigree information to monitor changes in the genetic diversity of captive populations, which is crucial to implement optimal mating decisions and assuring their long-term viability within an ex situ conservation programme.

Multiple and subject-specific roles of uncertainty in reward-guided decision-making.

Decision-making in noisy, changing, and partially observable environments entails a basic tradeoff between immediate reward and longer-term information gain, known as the exploration-exploitation dilemma. Computationally, an effective way to balance this tradeoff is by leveraging uncertainty to guide exploration. Yet, in humans, empirical findings are mixed, from suggesting uncertainty-seeking to indifference and avoidance. In a novel bandit task that better captures uncertainty-driven behavior, we find multiple roles for uncertainty in human choices. First, stable and psychologically meaningful individual differences in uncertainty preferences actually range from seeking to avoidance, which can manifest as null group-level effects. Second, uncertainty modulates the use of basic decision heuristics that imperfectly exploit immediate rewards: a repetition bias and win-stay-lose-shift heuristic. These heuristics interact with uncertainty, favoring heuristic choices under higher uncertainty. These results, highlighting the rich and varied structure of reward-based choice, are a step to understanding its functional basis and dysfunction in psychopathology.

Penetration State Recognition during Laser Welding Process Control Based on Two-Stage Temporal Convolutional Networks.

Vision-based laser penetration control has become an important research area in the field of welding quality control. Due to the complexity and large number of parameters in the monitoring model, control of the welding process based on deep learning and the reliance on long-term information for penetration identification are challenges. In this study, a penetration recognition method based on a two-stage temporal convolutional network is proposed to realize the online process control of laser welding. In this paper, a coaxial vision welding monitoring system is built. A lightweight segmentation model, based on channel pruning, is proposed to extract the key features of the molten pool and the keyhole from the clear molten pool keyhole image. Using these molten pool and keyhole features, a temporal convolutional network based on attention mechanism is established. The recognition method can effectively predict the laser welding penetration state, which depends on long-term information. In addition, the penetration identification experiment and closed-loop control experiment of unequal thickness plates are designed. The proposed method in this study has an accuracy of 98.96% and an average inference speed of 20.4 ms. The experimental results demonstrate that the proposed method exhibits significant performance in recognizing the penetration state from long sequences of welding image signals, adjusting welding power, and stabilizing welding quality.