The discovery rate of optical transients will explode to 10 million public alerts per night once the Vera C. Rubin Observatory’s Legacy Survey of Space and Time comes online, overwhelming the traditional physics-based inference pipelines. A continuous-time forecasting AI model is of interest because it can deliver millisecond-scale inference for thousands of objects per day, whereas legacy MCMC codes need hours per object. In this paper, we propose SELDON, a new continuous-time variational autoencoder for panels of sparse and irregularly time-sampled (gappy) astrophysical light curves that are nonstationary, heteroscedastic, and inherently dependent. SELDON combines a masked GRU-ODE encoder with a latent neural ODE propagator and an interpretable Gaussian-basis decoder. The encoder learns to summarize panels of imbalanced and correlated data even when only a handful of points are observed. The neural ODE then integrates this hidden state forward in continuous time, extrapolating to future unseen epochs. This extrapolated time series is further encoded by deep sets to a latent distribution that is decoded to a weighted sum of Gaussian basis functions, the parameters of which are physically meaningful. Such parameters (e.g., rise time, decay rate, peak flux) directly drive downstream prioritization of spectroscopic follow-up for astrophysical surveys. Beyond astronomy, the architecture of SELDON offers a generic recipe for interpretable and continuous-time sequence modeling in any time domain where data are multivariate, sparse, heteroscedastic, and irregularly spaced.

This paper presents a new approach for Gaussian process (GP) regression for large datasets. The approach involves partitioning the regression input domain into multiple local regions with a different local GP model fitted in each region. Unlike existing local partitioned GP approaches, we introduce a technique for patching together the local GP models nearly seamlessly to ensure that the local GP models for two neighboring regions produce nearly the same response prediction and prediction error variance on the boundary between the two regions. This largely mitigates the well-known discontinuity problem that degrades the prediction accuracy of existing local partitioned GP methods over regional boundaries. Our main innovation is to represent the continuity conditions as additional pseudo-observations that the differences between neighboring GP responses are identically zero at an appropriately chosen set of boundary input locations. To predict the response at any input location, we simply augment the actual response observations with the pseudo-observations and apply standard GP prediction methods to the augmented data. In contrast to heuristic continuity adjustments, this has an advantage of working within a formal GP framework, so that the GP-based predictive uncertainty quantification remains valid. Our approach also inherits a sparse block-like structure for the sample covariance matrix, which results in computationally efficient closed-form expressions for the predictive mean and variance. In addition, we provide a new spatial partitioning scheme based on a recursive space partitioning along local principal component directions, which makes the proposed approach applicable for regression domains having more than two dimensions. Using three spatial datasets and three higher dimensional datasets, we investigate the numerical performance of the approach and compare it to several state-of-the-art approaches.

Autoassociative neural networks (ANNs) have been proposed as a nonlinear extension of principal component analysis (PCA), which is commonly used to identify linear variation patterns in high-dimensional data. While principal component scores represent uncorrelated features, standard backpropagation methods for training ANNs provide no guarantee of producing distinct features, which is important for interpretability and for discovering the nature of the variation patterns in the data. Here, we present an alternating nonlinear PCA method, which encourages learning of distinct features in ANNs. A new measure motivated by the condition of orthogonal loadings in PCA is proposed for measuring the extent to which the nonlinear principal components represent distinct variation patterns. We demonstrate the effectiveness of our method using a simulated point cloud data set as well as a subset of the MNIST handwritten digits data. The results show that standard ANNs consistently mix the true variation sources in the low-dimensional representation learned by the model, whereas our alternating method produces solutions where the patterns are better separated in the low-dimensional space.

BACKGROUND AND OBJECTIVE: Electronic medical record (EMR) databases offer significant potential for developing clinical hypotheses and identifying disease risk associations by fitting statistical models that capture the relationship between a binary response variable and a set of predictor variables that represent clinical, phenotypical, and demographic data for the patient. However, EMR response data may be error prone for a variety of reasons. Performing a manual chart review to validate data accuracy is time consuming, which limits the number of chart reviews in a large database. The authors' objective is to develop a new design-of-experiments-based systematic chart validation and review (DSCVR) approach that is more powerful than the random validation sampling used in existing approaches. METHODS: The DSCVR approach judiciously and efficiently selects the cases to validate (i.e., validate whether the response values are correct for those cases) for maximum information content, based only on their predictor variable values. The final predictive model will be fit using only the validation sample, ignoring the remainder of the unvalidated and unreliable error-prone data. A Fisher information based D-optimality criterion is used, and an algorithm for optimizing it is developed. RESULTS: The authors' method is tested in a simulation comparison that is based on a sudden cardiac arrest case study with 23 041 patients' records. This DSCVR approach, using the Fisher information based D-optimality criterion, results in a fitted model with much better predictive performance, as measured by the receiver operating characteristic curve and the accuracy in predicting whether a patient will experience the event, than a model fitted using a random validation sample. CONCLUSIONS: The simulation comparisons demonstrate that this DSCVR approach can produce predictive models that are significantly better than those produced from random validation sampling, especially when the event rate is low.

Kernel principal component analysis (KPCA) is a method widely used for denoising multivariate data. Using geometric arguments, we investigate why a projection operation inherent to all existing KPCA denoising algorithms can sometimes cause very poor denoising. Based on this, we propose a modification to the projection operation that remedies this problem and can be incorporated into any of the existing KPCA algorithms. Using toy examples and real datasets, we show that the proposed algorithm can substantially improve denoising performance and is more robust to misspecification of an important tuning parameter.

Variance component estimation algorithms, in conjunction with automated in-process measurement technology, can be effective tools for identifying and eliminating major sources of manufacturing variation in assembly processes. Whether a particular set of variation sources are diagnosable depends critically on how the sensor system is laid out. Diagnosability tests are mathematical in nature and provide little insight into why a particular sensor layout may be nondiagnosable or how to modify the layout to ensure diagnosability. This paper translates the mathematical diagnosability conditions into a set of more conceptually meaningful conditions that provide better insight into the reasons behind the nondiagnosability. Note to Practitioners - This paper was motivated by the problem of identifying and eliminating major sources of variation in discrete-part manufacturing, which are critical steps in improving product quality. The effectiveness of statistical algorithms for estimating sources of variation depends on whether the sensor system for measuring key product and process variables is laid out properly, so that a particular set of diagnosability conditions are satisfied. This paper translates the rather abstract mathematical conditions for diagnosability into a set of more intuitive and conceptually meaningful conditions. This provides practitioners with insight into why a sensor system may be nondiagnosable and how to add or adjust sensors in order to ensure diagnosability. The diagnosability characterization can also be used to enhance performance and reduce computational expense in numerical search strategies for optimizing sensor layout.