Real-time Predictors Research Articles

Dynamically predicting comprehension difficulties through physiological data and intelligent wearables

Comprehending digital content written in natural language online is vital for many aspects of life, including learning, professional tasks, and decision-making. However, facing comprehension difficulties can have negative consequences for learning outcomes, critical thinking skills, decision-making, error rate, and productivity. This paper introduces an innovative approach to predict comprehension difficulties at the local content level (e.g., paragraphs). Using affordable wearable devices, we acquire physiological responses non-intrusively from the autonomous nervous system, specifically pulse rate variability, and electrodermal activity. Additionally, we integrate data from a cost-effective eye-tracker. Our machine learning algorithms identify ’hotspots’ within the content and regions corresponding to a high cognitive load. These hotspots represent real-time predictors of comprehension difficulties. By integrating physiological data with contextual information (such as the levels of experience of individuals), our approach achieves an accuracy of 72.11% ± 2.21, a precision of 0.77, a recall of 0.70, and an f1 score of 0.73. This study opens possibilities for developing intelligent, cognitive-aware interfaces. Such interfaces can provide immediate contextual support, mitigating comprehension challenges within content. Whether through translation, content generation, or content summarization using available Large Language Models, this approach has the potential to enhance language comprehension.

Real-time predictors of food parenting practices and child eating behaviors in racially/ethnically diverse families

BackgroundPrior research has shown associations between controlling food parenting practices (e.g., pressure-to-eat, restriction) and factors that increase risk for cardiovascular disease in children (e.g., low diet quality, obesity). This study aimed to examine associations between real-time parental stress and depressed mood, food parenting practices, and child eating behaviors in a longitudinal cohort study.MethodsChildren ages 5–9 years and their families (n = 631) from six racial/ethnic groups (African American, Hispanic, Hmong, Native American, Somali/Ethiopian, White) were recruited for this study through primary care clinics in a large metromolitan area in the US (Minneapolis/St. Paul, MN) in 2016–2019. Ecological momentary assessment was carried out over seven days with parents at two time points, 18 months apart. Adjusted associations between morning stress and depressed mood of parents on food parenting practices and child eating behaviors at the evening meal were examined. Interactions tested whether food security, race/ethnicity and child sex moderated associations.ResultsHigh levels of parental stress and depressed mood experienced earlier in the day were associated with controlling food parenting practices and child food fussiness at dinner the same night. Results were dependent on food security status, race/ethnicity, and child sex.ConclusionsHealth care professionals may want to consider, or continue, screening parents for stress, depression, and food insecurity during well-child visits and discuss the influence these factors may have on food parenting practices and child eating behaviors. Future research should use real-time interventions such as ecological momentary intervention to reduce parental stress and depressed mood to promote healthy food parenting practices and child eating behaviors.

Use of Real-Time Information to Predict Future Arrivals in the Emergency Department.

We aimed to build prediction models for shift-level emergency department (ED) patient volume that could be used to facilitate prediction-driven staffing. We sought to evaluate the predictive power of rich real-time information and understand 1) which real-time information had predictive power and 2) what prediction techniques were appropriate for forecasting ED demand. We conducted a retrospective study in an ED site in a large academic hospital in New York City. We examined various prediction techniques, including linear regression, regression trees, extreme gradient boosting, and time series models. By comparing models with and without real-time predictors, we assessed the potential gain in prediction accuracy from real-time information. Real-time predictors improved prediction accuracy on models without contemporary information from 5% to 11%. Among extensive real-time predictors examined, recent patient arrival counts, weather, Google trends, and concurrent patient comorbidity information had significant predictive power. Out of all the forecasting techniques explored, SARIMAX (Seasonal Autoregressive Integrated Moving Average with eXogenous factors) achieved the smallest out-of-sample the root mean square error (RMSE) of 14.656 and mean absolute prediction error (MAPE) of 8.703%. Linear regression was the second best, with out-of-sample RMSE and MAPE equal to 15.366 and 9.109%, respectively. Real-time information was effective in improving the prediction accuracy of ED demand. Practice and policy implications for designing staffing paradigms with real-time demand forecasts to reduce ED congestion were discussed.

Real-time momentary mood as a predictor of college students' prescription drug misuse in daily life: Direct links and the moderating role of background mental health.

The present study addressed calls for research to identify real-time predictors of prescription drug misuse (Schepis et al., 2020) by testing young adults' momentary reports of their negative mood and positive mood as predictors of event-level misuse in daily life. We implemented a 28-day ecological momentary assessment (EMA) procedure that collected individuals' mood and other contextual experiences in moments preceding prescription drug misuse. Consistent with models of problematic substance use as a means to reduce negativity (Khantzian, 1997), results from hierarchical generalized linear modeling (HGLM) indicated within-person links between higher than usual negative mood and greater likelihood of prescription misuse in daily life. Contrary to the hypothesis, misuse was also more likely when preceded by elevated positive mood. We found consistent support for the hypothesized between-person effects, with prescription misuse in daily life associated with higher average levels of negative mood, and lower average levels of positive mood, across the reporting period. We further predicted that individuals reporting greater levels of social anxiety, depression, and externalizing symptoms would evidence stronger links between their momentary negative mood and prescription misuse. Partial support for this moderation hypothesis was found, with the positive within-person link between negative mood and prescription misuse significantly stronger among individuals higher (vs. lower) on social anxiety and depression. Results provide support for intricate connections between young adults' momentary mood, mental health symptoms, and prescription drug misuse. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Use of Random Domain Intercept Technology to Track COVID-19 Vaccination Rates in Real Time Across the United States: Survey Study.

BackgroundAccurate and timely COVID-19 vaccination coverage data are vital for informing targeted, effective messaging and outreach and identifying barriers to equitable health service access. However, gathering vaccination rate data is challenging, and efforts often result in information that is either limited in scope (eg, limited to administrative data) or delayed (impeding the ability to rapidly respond). The evaluation of innovative technologies and approaches that can assist in addressing these limitations globally are needed.ObjectiveThe objective of this survey study was to assess the validity of Random Domain Intercept Technology (RDIT; RIWI Corp) for tracking self-reported vaccination rates in real time at the US national and state levels. RDIT—a form of online intercept sampling—has the potential to address the limitations of current vaccination tracking systems by allowing for the measurement of additional data (eg, attitudinal data) and real-time, rapid data collection anywhere there is web access.MethodsWe used RDIT from June 30 to July 26, 2021, to reach a broad sample of US adult (aged ≥18 years) web users and asked questions related to COVID-19 vaccination. Self-reported vaccination status was used as the focus of this validation exercise. National- and state-level RDIT-based vaccination rates were compared to Centers for Disease Control and Prevention (CDC)–reported national and state vaccination rates. Johns Hopkins University’s and Emory University’s institutional review boards designated this project as public health practice to inform message development (not human subjects research).ResultsBy using RDIT, 63,853 adult web users reported their vaccination status (6.2% of the entire 1,026,850 American web-using population that was exposed to the survey). At the national level, the RDIT-based estimate of adult COVID-19 vaccine coverage was slightly higher (44,524/63,853, 69.7%; 95% CI 69.4%-70.1%) than the CDC-reported estimate (67.9%) on July 15, 2021 (ie, midway through data collection; t63,852=10.06; P<.001). The RDIT-based and CDC-reported state-level estimates were strongly and positively correlated (r=0.90; P<.001). RDIT-based estimates were within 5 percentage points of the CDC’s estimates for 29 states.ConclusionsThis broad-reaching, real-time data stream may provide unique advantages for tracking the use of a range of vaccines and for the timely evaluation of vaccination interventions. Moreover, RDIT could be harnessed to rapidly assess demographic, attitudinal, and behavioral constructs that are not available in administrative data, which could allow for deeper insights into the real-time predictors of vaccine uptake–enabling targeted and timely interventions.

Opening the Black Box of Daily Life in Nonsuicidal Self-injury Research: With Great Opportunity Comes Great Responsibility.

Although nonsuicidal self-injury (NSSI)—deliberate damaging of body tissue without suicidal intent—is a behavior that occurs in interaction with real-world contexts, studying NSSI in the natural environment has historically been impossible. Recent advances in real-time monitoring technologies have revolutionized our ability to do exactly that, providing myriad research and clinical practice opportunities. In this viewpoint paper, we review new research pathways to improve our ability to understand, predict, and prevent NSSI, and provide critical perspectives on the responsibilities inherent to conducting real-time monitoring studies on NSSI. Real-time monitoring brings unique opportunities to advance scientific understanding about (1) the dynamic course of NSSI, (2) the real-time predictors thereof and ability to detect acute risk, (3) the ecological validity of theoretical models, (4) the functional mechanisms and outcomes of NSSI, and (5) the promotion of person-centered care and novel technology-based interventions. By considering the opportunities of real-time monitoring research in the context of the accompanying responsibilities (eg, inclusive recruitment, sound and transparent research practices, participant safety and engagement, measurement reactivity, researcher well-being and training), we provide novel insights and resources to open the black box of daily life in the next decade(s) of NSSI research.

Real-time predictors of body dissatisfaction in females with binge eating: an ecological momentary assessment study.

Body dissatisfaction (BD) is supported as a maintenance factor for eating disorders (EDs) characterized by binge eating (BE). Although it is traditionally conceptualized as a trait construct, ecological momentary assessment (EMA) studies have shown that it fluctuates within-day and that momentary elevations in BD prospectively predict near-time ED behaviors. However, less is known about the contextual factors that precipitate these fluctuations in BD. The current study thus sought to examine whether certain internal (i.e., negative affect, shape/weight-related cognitions) and environmental (i.e., attending social events, being exposed to thin-ideal media, spending time on social media) factors prospectively predict momentary elevations in BD in females with BE. Participants (n = 25) completed an EMA protocol over 14 days. Data on BD and internal and environmental factors were analyzed using generalized estimating equations. Results showed that (1) greater than one's usual negative affect and shape/weight-related cognitions, and (2) spending time on social media prospectively predicted momentary elevations in BD. Interestingly, attending social events prospectively predicted momentary reductions in BD. These findings may have important implications for conceptualizing and managing BD to prevent ED behaviors. Level II, controlled trial without randomization.

Rationale and Design of the Women's Health And Daily Experiences Project: Protocol for an Ecological Momentary Assessment Study to Identify Real-Time Predictors of Midlife Women's Physical Activity.

BackgroundMidlife women are at an elevated risk for cardiovascular disease (CVD) and associated mortality. Those who have additional risk conditions such as obesity or hypertension report specific barriers to engaging in cardioprotective behaviors such as physical activity (PA). Considerable effort has been devoted to understanding PA determinants and designing interventions for midlife women, although with suboptimal success, as increasing PA could meaningfully attenuate CVD risk. An updated approach to understanding PA among midlife women could improve upon existing resources by focusing on novel psychosocial influences on PA in this population (ie, body satisfaction, social interactions, social comparisons, mood state) and within-person relations between these influences and PA in the natural environment.ObjectiveThe overarching goal of Project WHADE (Women’s Health And Daily Experiences) is to use an ecological momentary assessment (EMA) approach to capture ecologically valid relations between midlife women’s psychosocial experiences and PA as they engage in their normal daily activities. The primary aim of the study is to identify within-person psychosocial predictors of variability in PA (ie, experiences associated with higher vs lower PA for a given individual).MethodsMidlife women (aged 40-60 years) with one or more additional risk markers for CVD (eg, hypertension) will be recruited from primary care clinics and the general community (target n=100). Eligible women will complete an initial survey and a face-to-face baseline session before engaging in a 10-day EMA protocol. Psychosocial experiences will be assessed using a brief self-report via a smartphone 5 times per day, and PA will be assessed throughout waking hours using a research-grade monitor. Participants will return for a brief exit interview at the end of 10 days. Multilevel models that address the nested structure of EMA data will be used to evaluate the study aims.ResultsRecruitment and enrollment are ongoing, and a total of 75 women have completed the protocol to date. Data collection is expected to be completed in Fall 2020.ConclusionsProject WHADE is designed to identify naturally occurring psychosocial experiences that predict short-term variability in midlife women’s PA. As such, the results of this study should advance the current understanding of PA among midlife women by providing further insight into within-person psychosocial influences on PA in this group. In the future, this information could help inform the design of interventions for this population.International Registered Report Identifier (IRRID)DERR1-10.2196/19044

The utility of blood-based molecular tools-the NETest-to monitor and evaluate the efficacy of PRRT in neuroendocrine tumors.

3568 Background: Peptide receptor radionuclide therapy (PRRT) is an effective therapy for metastatic/inoperable neuroendocrine tumors (NETs). Tools to predict and monitor the efficacy of therapy are important adjuncts in the radio-oncology armamentarium. Standard blood biomarkers are not effective by new molecular based assays such as the PRRT Predictive Quotient (PPQ) and NETest are effective as real-time predictors and monitors of therapy. We aimed to prospectively evaluate whether: 1) the NETest functioned as a surrogate biomarker for image-based per RECIST evaluation of PRRT efficacy; 2) there was a correlation between changes in NETest levels during therapy, PPQ prediction and treatment efficacy. Methods: Three independent 177Lu-PRRT-treated GEP-NET and BPNEN cohorts (Rotterdam, Netherlands: n= 41; Bad-Berka, Germany: n= 44; Meldola, Italy: n= 72). Treatment response: RECIST1.1 [Responder (stable, partial/complete response) vs Non-Responder]. Blood sampling: pre-PRRT, prior to each cycle and 6 months (median) after completion of all cycles. PPQ (positive/negative) and NETest (0-100 score) by PCR. Stable&lt;40; progressive &gt; 40). CgA (ELISA) as comparator. Samples deidentified, measurement and analyses blinded. Kaplan-Meier survival and Mann-Whitney analyses. Results: 122 of 157 were evaluable. RECIST stabilization or response in 67%; 33% progressed. NETest significantly ( p&lt; 0.0001) decreased in RECIST-“responders” (-47±3%); in “non-responders” it elevated (+79±19%, p&lt; 0.0005). NETest monitoring accuracy 98% (119/122). Follow-up levels &gt; 40 (progressive) vs stable (&lt;40) significantly correlated with mPFS (not reached vs. 10 months; HR 0.04, 95%CI: 0.02-0.07). PPQ response prediction was accurate in 118 (97%); 99% accurate positive and 93% accurate negative prediction. NETest significantly ( p&lt; 0.0001) decreased in PPQ-predicted responders (-46±3%) and remained increased in PPQ-predicted non-responders (+75±19%). Follow-up NETest categories stable vs progressive significantly correlated with PPQ prediction and mPFS (not reached vs. 10 months; HR 0.06, 95%CI: 0.03-0.12). In comparison, the standard biomarker, CgA, failed to predict or correlate with response to PRRT ( p= NS). Conclusions: NETest accurately (98%) monitors PRRT response and is an effective surrogate marker for radiological response (image concordance 98%). A NETest decrease identified responders (99%) and correlated ( &gt; 97%) with the pretreatment PPQ response predictor.

Fluctuations in Affective States and Self-Efficacy to Resist Non-Suicidal Self-Injury as Real-Time Predictors of Non-Suicidal Self-Injurious Thoughts and Behaviors.

IntroductionAlthough research over the past decade has resulted in significantly increased knowledge about distal risk factors for non-suicidal self-injury (NSSI), little is known about short-term (proximal) factors that predict NSSI thoughts and behaviors. Drawing on contemporaneous theories of NSSI, as well as the concept of ideation-to-action, the present study clarifies (a) real-time factors that predict NSSI thoughts and (b) the extent to which theoretically important momentary factors (i.e., negative affect, positive affect, and self-efficacy to resist NSSI) predict NSSI behavior in daily life, beyond NSSI thoughts.MethodsUsing experience sampling methodology, intensive longitudinal data was obtained from 30 young adults with frequent NSSI episodes in the last year. Participants completed assessments up to eight times per day for 12 consecutive days (signal-contingent sampling). This resulted in the collection of 2,222 assessments (median compliance = 79.2%) during which 591 NSSI thoughts and 270 NSSI behaviors were recorded. Using the dynamic structural equation modeling framework, multilevel vector autoregressive models were constructed.ResultsWithin the same assessment, negative affect was positively associated with NSSI thoughts, whereas positive affect and self-efficacy to resist NSSI were each negatively associated with NSSI thoughts. Across assessments, higher-than-usual negative affect and self-efficacy to resist NSSI were predictive of short-term change in NSSI thoughts. While fluctuations in both negative affect and positive affect prospectively predicted NSSI behavior, these factors became non-significant in models that controlled for the predictive effect of NSSI thoughts. In contrast, self-efficacy to resist NSSI incrementally predicted a lower probability of engaging in NSSI, above and beyond NSSI thoughts.DiscussionThis study provides preliminary evidence that affective fluctuations may uniquely predict NSSI thoughts but not NSSI behaviors, and point to the role of personal belief in the ability to resist NSSI in preventing NSSI behavior. These findings illustrate the need to differentiate between the development of NSSI thoughts and the progression from NSSI thoughts to behavior, as these are likely distinct processes, with different predictors.

Real-time predictors and consequences of binge eating among adults with type 1 diabetes

BackgroundObjective binge eating (OBE) is common among individuals with type 1 diabetes (T1D) and may have negative consequences for glycemic control. Recent studies have suggested that diabetes distress (i.e., emotional distress specific to diabetes and living with the burden of management) is a distinct emotional experience among individuals with diabetes. Preliminary studies have found diabetes distress is associated with eating disorder symptoms and poor glycemic control. The aim of the current study was to examine real-time emotional precursors and consequences of OBE in adults with T1D (i.e., general negative affect, specific emotional states and diabetes distress) using ecological momentary assessment methods. We also explore the impact of OBE on 2-h postprandial glycemic control relative to non-OBE eating episodes.MethodsAdults with T1D (N = 83) completed 3-days of ecological momentary assessment assessing mood and eating behavior using a telephone-based survey system. Participants were prompted to rate momentary affect, including level diabetes distress, at random intervals and reported on eating episodes. Participants also wore continuous glucose monitors allowing for ongoing assessment of glycemic control. Multi-level modeling was used to examine between- and within-person effects of momentary increases in emotions prior to eating on the likelihood of OBE and the impact of OBE on postprandial blood glucose. Generalized linear mixed models examined whether change in post-meal affect differed between OBE and non-OBE episodes.ResultsParticipants were predominately middle-aged (Mean = 42; SD = 12.43) Caucasian (87%) females (88%) reporting clinically significant eating disorder symptoms (76%). Nearly half of the sample (43%) reported OBE during the 3-day study period. The between-person effect for negative affect was significant (OR = 1.93, p < .05), indicating a 93% increased risk of OBE among individuals with higher negative affect compared to individuals with average negative affect. Between-person effects were also significant for guilt, frustration and diabetes distress (OR = 1.48–1.77, ps < .05). Analyses indicated that mean change in post-meal negative affect was significantly greater for OBE relative to non-OBE episodes (B = 0.44, p < .001). Blood glucose at 120 min postprandial was also higher for OBE than for non-OBE episodes (p = .03).ConclusionsFindings indicate that individuals who tend to experience negative affect and diabetes distress before eating are at increased risk of OBE at the upcoming meal. Results also suggest that engaging in binge eating may result in greater subsequent negative affect, including diabetes distress, and lead to elevated postprandial blood glucose levels. These findings add to a growing literature suggesting diabetes distress is related to eating disordered behaviors among individuals with T1D.

Recurrent neural networks for real-time prediction of TBM operating parameters

Abstract With tunnel boring machines (TBMs) widely used in tunnel construction, the adaptable adjustment of TBM operating status has become a research focus. Since the prediction of tunnel geological conditions is still challenging before excavation, the prediction of important TBM operating parameters plays an important role in the research on TBM adaptable adjustment. In this paper, we use three kinds of recurrent neural networks (RNNs), including traditional RNNs, long-short term memory (LSTM) networks and gated recurrent unit (GRU) networks, to deal with the real-time prediction of TBM operating parameters based on TBM in-situ operating data. The experimental results show that the proposed three kinds of RNN-based predictors can provide accurate prediction values of some important TBM operating parameters during next period including the torque, the velocity, the thrust and the chamber pressure. We also make a comparison with several classical regression models (e.g., support vector regression (SVR), random forest (RF) and Lasso) which actually cannot act as real-time predictors in a real sense, and the comparative experiments show that the proposed RNN-based predictors outperform the regression models in most cases. The feasibility of RNNs for the real-time prediction of TBM operating parameters indicates that RNNs can afford the analysis and the forecasting of the time-continuous in-situ data collected from various construction equipments.

Real-time predictors of smoking among sexual minority and heterosexual young adults: An ecological momentary assessment study

BackgroundSexual minority young adults have higher smoking rates than the general young adult population, but reasons for this disparity are poorly understood. The current study aimed to: 1) identify real-time predictors of smoking among sexual minority and heterosexual smokers and 2) examine between-group differences in these predictors. MethodsWe conducted an ecological momentary assessment (EMA) study in the San Francisco Bay Area, California in 2016-2017. Data from 84 young adult smokers (44% identified as sexual minority, including 29 bisexual and 8 gay/lesbian) with 6498 EMA assessments were analyzed. Both internal and external predictors and interaction terms between each predictor and sexual group were examined using generalized estimating equation models. ResultsCommon correlates of smoking were found for both groups (e.g., craving, absence of smoking bans, presence of other smokers, outside location, and seeing triggers). Unique factors for sexual minority smokers were being at a bar (aOR = 1.75, 95% CI = 1.06–2.90) and the number of other smokers present (aOR = 1.12, 95%CI = 1.04–1.20), while the presence of a smoking family member reduced the odds of smoking in this group (aOR = 0.13, 95%CI = 0.02-0.85). In interaction models, the number of other smokers exerted a greater influence on sexual minority participants compared to their heterosexual counterparts (aOR = 1.10, 95%CI = 1.01–1.20), while craving (aOR = 0.84, 95%CI = 0.75-0.93) and presence of a smoking family member (aOR = 0.11, 95%CI = 0.01-0.82) had weaker influences. ConclusionsOur study highlights unique situational factors associated with smoking among sexual minority young adults and differences in these factors by sexual identity. Future interventions targeting sexual minorities should address bar attendance and specific triggers.

A negative group delay model for feedback-delayed manual tracking performance.

We propose that feedback-delayed manual tracking performance is limited by fundamental constraints imposed by the physics of negative group delay. To test this hypothesis, the results of an experiment in which subjects demonstrate both reactive and predictive dynamics are modeled by a linear system with delay-induced negative group delay. Although one of the simplest real-time predictors conceivable, this model explains key components of experimental observations. Most notably, it explains the observation that prediction time linearly increases with feedback delay, up to a certain point when tracking performance deteriorates. It also explains the transition from reactive to predictive behavior with increasing feedback delay. The model contains only one free parameter, the feedback gain, which has been fixed by comparison with one set of experimental observations for the reactive case. Our model provides quantitative predictions that can be tested in further experiments.

Predicting wake meandering in real-time through instantaneous measurements of wind turbine load fluctuations

The present study deals with the real-time estimation of the wind turbine wake meandering and the determination of the best indicators to measure in order to build a real-time wake meandering model. Results are obtained through experiments performed in an atmospheric boundary layer wind tunnel, where a specific set-up enables to measure simultaneously the incoming lateral velocity fluctuations, the lateral force fluctuations applied to the wind turbine model and to track in real-time the lateral position of its wake. It is an extension of a previous work about the determination of good candidates to build some real-time predictors of the wake meandering. The strong correlations between the incoming transverse velocity fluctuations, the lateral force fluctuations and the lateral position of the wake farther downstream are quantified, confirming that the large-scale turbulent eddies impact directly the wake meandering. Three different versions of a wake meandering predictor model are compared. It leads to the conclusion that the monitoring of the global force fluctuations applied to a wind turbine could be used to predict in real-time the meandering of the generated wake.

TH-AB-303-03: Real-Time Error Estimation for Real-Time Motion Prediction

Purpose: A critical component of real-time MLC tracking is to incorporate motion prediction in order to account for system latency, the time between the occurrence of motion and MLC leaf repositioning, so as to avoid geometric and dosimetric errors. In this work, we address a major limitation of current motion prediction algorithms – estimating the inherent error in prediction in real-time. We propose using an ensemble of real-time predictors to reduce overall prediction error, and provide a quantitative metric to trigger beam holds based on a user-defined error threshold. Methods: An ensemble of eighteen kernel density estimation (KDE)-based estimators was implemented in parallel and deployed on motion traces recorded from ten lung cancer patients. For each trace, the ensemble generated a sequence of predictions for a look-ahead time of 250ms, typical for many real-time systems. Each predictor, and the ensemble as a whole, was evaluated based on their root-mean-square error (RMSE) with respect to the original data, and assigned a beam hold score between 0 (perfect) and 1 to quantify the number of false beam holds and false beam releases. Results: The average execution time of the KDE-ensemble was 16ms, well below the 250ms look-ahead. For all traces, the ensemble produced a lower RMSE and hold score than any of the individual constituent predictors. For the worst-case motion trace, the minimum RMSE(hold score) for individual predictors was 0.912cm(0.340) compared to 0.854cm(0.305) for the KDE-ensemble predictor. The hold score for individual estimators was ∼0.6 when error estimates were not used to assert beam holds. Conclusion: The KDE-ensemble approach results in improved motion prediction. More importantly, it also provides real-time estimations of prediction error, thus addressing a major gap in current motion prediction algorithms. Such error prediction can be utilized to assert beam-holds during MLC tracking, preventing serious geometric and dosimetric errors. Funding for this work was provided by Elekta Limited.

Determination of real-time predictors of the wind turbine wake meandering

The present work proposes an experimental methodology to characterize the unsteady properties of a wind turbine wake, called meandering, and particularly its ability to follow the large-scale motions induced by large turbulent eddies contained in the approach flow. The measurements were made in an atmospheric boundary layer wind tunnel. The wind turbine model is based on the actuator disc concept. One part of the work has been dedicated to the development of a methodology for horizontal wake tracking by mean of a transverse hot wire rake, whose dynamic response is adequate for spectral analysis. Spectral coherence analysis shows that the horizontal position of the wake correlates well with the upstream transverse velocity, especially for wavelength larger than three times the diameter of the disc but less so for smaller scales. Therefore, it is concluded that the wake is actually a rather passive tracer of the large surrounding turbulent structures. The influence of the rotor size and downstream distance on the wake meandering is studied. The fluctuations of the lateral force and the yawing torque affecting the wind turbine model are also measured and correlated with the wake meandering. Two approach flow configurations are then tested: an undisturbed incoming flow (modelled atmospheric boundary layer) and a disturbed incoming flow, with a wind turbine model located upstream. Results showed that the meandering process is amplified by the presence of the upstream wake. It is shown that the coherence between the lateral force fluctuations and the horizontal wake position is significant up to length scales larger than twice the wind turbine model diameter. This leads to the conclusion that the lateral force is a better candidate than the upstream transverse velocity to predict in real time the meandering process, for either undisturbed (wake free) or disturbed incoming atmospheric flows.

Investigating Real-time Predictors of Engagement

Engagement is a worthwhile psychological construct to examine in the context of video games and online training. In this context, previous research suggests that the more engaged a person is, the more likely they are to experience overall positive affect while performing at a high level. This research builds on theories of engagement, Flow Theory, and Cognitive Load Theory, to operationalize engagement in terms of cognitive load, affect, and performance. An adaptive algorithm was then developed to test the proposed operationalization of engagement. Using a puzzle-based video game, player performance and engagement was compared across three conditions: adaptive gameplay, a traditional linear gameplay, and choice-based gameplay. Results show that those in the adaptive gameplay condition performed higher compared to those in the other two conditions without any degradation of overall affect or self-report of engagement.

Development of an efficient real-time disruption predictor from scratch on JET and implications for ITER

Prediction of disruptions from scratch is an ITER-relevant topic. The first operations with the new ITER-like wall constitute a good opportunity to test the development of new predictors from scratch and the related methodologies. These methodologies have been based on the Advanced Predictor Of DISruptions (APODIS) architecture. APODIS is a real-time disruption predictor that is in operation in the JET real-time network. Balanced and unbalanced datasets are used to develop real-time predictors from scratch. The discharges are used in chronological order. Also, different criteria to decide when to re-train a predictor are discussed. The best results are obtained by applying a hybrid method (balanced/unbalanced datasets) for training and with the criterion of re-training after every missed alarm. The predictors are tested off-line with all the discharges (disruptive/non-disruptive) corresponding to the first three JET ITER-like wall campaigns. The results give a success rate of 93.8% and a false alarm rate of 2.8%. It should be considered that these results are obtained from models trained with no more than 42 disruptive discharges.

Real-time predictors of suicidal ideation: Mobile assessment of hospitalized depressed patients

Suicidal ideation is a risk factor for suicide attempt and completion. Cross-sectional or retrospective studies cannot capture the dynamic course and possible predictors of suicidal ideation as it occurs in daily life. This study utilizes an experience sampling paradigm to identify real-time predictors of suicidal ideation in inpatients with major depressive disorder. Thirty-one depressed patients admitted to a psychiatric unit were signaled by a mobile device to record suicidal ideation, affect, and other symptoms, multiple times a day over 1week. Participants completed a total of 1350 questionnaires. Seventy-four percent of the sample reported suicidal ideation during the week. Time-lagged analyses revealed that momentary ratings of Sadness, Tension, and Boredom (as well as suicidal ideation itself) predicted subsequent suicidal thoughts in the following hours. Baseline severity of depression and past suicide attempts were both correlated with mean ideation severity during the week. A number of predictors identified in prior research (e.g. hopelessness) were unrelated to subsequent suicidal ideation in the current study. Momentary interventions that guide individuals through activities designed to reduce levels of Sadness, Tension, and Boredom in real-time (e.g., thought challenging, relaxation, behavioral activation) may be especially warranted.