Rider University is a private university in Lawrence Township, New Jersey. It consists of four academic units: the Norm Brodsky College of Business, the College of Liberal Arts and Sciences, the College of Education and Human Services, and Westminster College of the Arts (consisting of the School of Fine and Performing Arts and Westminster Choir College).
Training a single network with multiple objectives often leads to conflicting gradients that degrade shared representations, forcing them into a compromised state that is suboptimal for any single task—a problem we term latent representation collapse. We introduce Domain Expansion, a framework that prevents these conflicts by restructuring the latent space itself. Our framework uses a novel orthogonal pooling to construct a latent space where each objective is assigned to a mutually orthogonal subspace. We validate our approach on the ShapeNet benchmark, simultaneously training a model for object classification and pose estimation. Our experiments demonstrate that this structure not only prevents collapse but also yields an explicit, interpretable, and compositional latent space where concepts can be directly manipulated.
Environmental time series, such as near-surface air temperature, exhibit strong multi-scale structure and persistent autocorrelation. Accurate forecasting therefore requires careful consideration of both temporal scale separation and serial dependence. In this study, we evaluate a unified framework that integrates Kolmogorov–Zurbenko (KZ) filtering with two classes of models: (i) classical regression with Cochrane–Orcutt autocorrelation correction, and (ii) an autocorrelation-adjusted Long Short-Term Memory (LSTM) network that learns an embedded correlation coefficient (ρ). All models are assessed using standardized meteorological predictors of T2M under walk-forward validation. The LSTM trained on raw predictors shows moderate performance (RMSE = 0.73, R2=0.46, DW = 0.79), which improves after KZ filtering (RMSE = 0.59, R2=0.63, DW = 1.84). Classical regression applied to KZ-decomposed predictors and corrected using the Cochrane–Orcutt procedure achieves substantially higher accuracy (RMSE = 0.41, R2=0.89, DW ≈2.0), outperforming the LSTM in both predictive precision and residual behavior. Visual diagnostics further confirm tighter predicted–actual alignment and near-white residuals in the classical models, whereas the LSTM retains small systematic deviations even after filtering. Overall, the results demonstrate that addressing multi-scale structures and autocorrelation had a greater impact than increasing model complexity. Integrating spectral decomposition with autocorrelation correction thus produces more reliable, statistically valid forecasts, demonstrating that classical regression with KZ filtering can surpass LSTM models in both accuracy and interpretability. These findings emphasize the value of combining time series–aware pre-processing with both traditional and neural network approaches for environmental prediction.
Probabilistic streamflow models play a pivotal role in quantifying hydrological uncertainty and form the backbone of modern risk management strategies for flood and drought forecasting, water allocation planning, and the design of resilient infrastructure. Unlike deterministic approaches that yield single-point estimates, these models provide a spectrum of possible outcomes, enabling a more realistic assessment of extreme events and supporting informed, sustainable water resource decisions. By explicitly accounting for natural variability and uncertainty, probabilistic models promote transparent, robust, and equitable risk evaluations, helping decision-makers balance economic costs, societal benefits, and environmental protection for long-term sustainability. In this study, we introduce the bounded half-logistic distribution (BHLD), a novel heavy-tailed probability model constructed using the T-Y method for distribution generation, where T denotes a transformer distribution and Y represents a baseline generator. Although the BHLD is conceptually related to the Pareto and log-logistic families, it offers several distinctive advantages for streamflow modeling, including a flexible hazard rate that can be unimodal or monotonically decreasing, a finite lower bound, and closed-form expressions for key risk measures such as Value at Risk (VaR) and Tail Value at Risk (TVaR). The proposed distribution is defined on a lower-bounded domain, allowing it to realistically capture physical constraints inherent in flood processes, while a log-logistic-based tail structure provides the flexibility needed to model extreme hydrological events. Moreover, the BHLD is analytically characterized through a governing differential equation and further examined via its characteristic function and the maximum entropy principle, ensuring stable and efficient parameter estimation. It integrates a half-logistic generator with a log-logistic baseline, yielding a power-law tail decay governed by the parameter beta, which is particularly effective for representing extreme flows. Fundamental properties, including the hazard rate function, moments, and entropy measures, are derived in closed form, and model parameters are estimated using the maximum likelihood method. Applied to four real streamflow data sets, the BHLD demonstrates superior performance over nine competing distributions in goodness-of-fit analyses, with notable improvements in tail representation. The model facilitates accurate computation of hydrological risk metrics such as VaR, TVaR, and tail variance, uncovering pronounced temporal variations in flood risk and establishing the BHLD as a powerful and reliable tool for streamflow modeling under changing environmental conditions.
The Process Oriented Guided Inquiry Learning for Physical Chemistry Laboratory (POGIL-PCL) community has developed guided-inquiry experiments for physical chemistry laboratory courses with the goal of improving student engagement and meaningful learning. The present study evaluated and compared the extent to which four POGIL-PCL experiments and four analogous traditional experiments support student engagement in scientific practices. To account for the specific context of POGIL-PCL, we developed a modified three-level version of the Three-Dimensional Learning Assessment Protocol that considers the role of instructor facilitation and the context of advanced students who have become integrated into their communities of practice. In comparison to their traditional counterparts, the POGIL-PCL experiments were found to offer more consistent and higher-level opportunities for student engagement in scientific practices. The POGIL-PCL experiments showed stronger alignment with some practices more than others, namely, analyzing and interpreting data, using mathematics and computational thinking, developing and using models, and planning and carrying out investigations. The cyclic nature of POGIL-PCL provided students with repeated and iterative opportunities for engagement with scientific practices within a given experiment, and it was found that those opportunities for engagement in certain scientific practices were concentrated in particular parts of the POGIL-PCL data-think cycle. These findings provide insight into how POGIL-PCL experiments can be strengthened by the incorporation of prompts that more deliberately cue for additional scientific practices as well as through revisions to make engagement with the practices more consistent across a variety of contexts.
Developmental disabilities are prevalent among U.S. children, child disability rates have been increasing, and the increases have been driven by cognitive and behavioral disorders. This study estimates the effects of low cognitive test scores and high behavior problem scores in childhood on educational attainment, employment, wages, and access to transportation and credit in adulthood. We assess cognitive and behavior scores at multiple time points during childhood and estimate cross-household and household fixed-effects models. We find that individuals with low cognitive scores in childhood are 10% less likely to graduate from high school, 23% less likely to be employed, 31% less likely to own a motor vehicle, and 18% less likely to have a credit card, and they have 51% lower earnings compared with individuals with higher cognitive scores. We also find that individuals with high behavior problem scores in childhood are 7% less likely to graduate from high school, 11% less likely to be employed, and 13% less likely to own a motor vehicle, and they have 14% lower earnings compared with those with lower behavior problem scores. The findings have important implications for well-being over the life course for a nontrivial share of the U.S. population as well as their families and communities.
