More Like this

1. Forecasting the Evolution of Hydropower Generation

   https://doi.org/10.1145/3394486.3403337  

   Zhou, Fan ; Li, Liang ; Zhang, Kunpeng ; Trajcevski, Goce ; Yao, Fuming ; Huang, Ying ; Zhong, Ting ; Wang, Jiahao ; Liu, Qiao ( July 2020 , The 26th {ACM} {SIGKDD} Conference on Knowledge Discovery and Data Mining, CA, August 2020)

   null (Ed.)

   Hydropower is the largest renewable energy source for electricity generation in the world, with numerous benefits in terms of: environment protection (near-zero air pollution and climate impact), cost-effectiveness (long-term use, without significant impacts of market fluctuation), and reliability (quickly respond to surge in demand). However, the effectiveness of hydropower plants is affected by multiple factors such as reservoir capacity, rainfall, temperature and fluctuating electricity demand, and particularly their complicated relationships, which make the prediction/recommendation of station operational output a difficult challenge. In this paper, we present DeepHydro, a novel stochastic method for modeling multivariate time series (e.g., water inflow/outflow and temperature) and forecasting power generation of hydropower stations. DeepHydro captures temporal dependencies in co-evolving time series with a new conditioned latent recurrent neural networks, which not only considers the hidden states of observations but also preserves the uncertainty of latent variables. We introduce a generative network parameterized on a continuous normalizing flow to approximate the complex posterior distribution of multivariate time series data, and further use neural ordinary differential equations to estimate the continuous-time dynamics of the latent variables constituting the observable data. This allows our model to deal with the discrete observations in the context of continuous dynamic systems, while being robust to the noise. We conduct extensive experiments on real-world datasets from a large power generation company consisting of cascade hydropower stations. The experimental results demonstrate that the proposed method can effectively predict the power production and significantly outperform the possible candidate baseline approaches. 

   more » « less
2. Enabling hyperparameter optimization in sequential autoencoders for spiking neural data

   Keshtkaran, MR ; Pandarinath, C ( January 2019 , Advances in Neural Information Processing Systems)

   Continuing advances in neural interfaces have enabled simultaneous monitoring of spiking activity from hundreds to thousands of neurons. To interpret these large-scale data, several methods have been proposed to infer latent dynamic structure from high-dimensional datasets. One recent line of work uses recurrent neural networks in a sequential autoencoder (SAE) framework to uncover dynamics. SAEs are an appealing option for modeling nonlinear dynamical systems, and enable a precise link between neural activity and behavior on a single-trial basis. However, the very large parameter count and complexity of SAEs relative to other models has caused concern that SAEs may only perform well on very large training sets. We hypothesized that with a method to systematically optimize hyperparameters (HPs), SAEs might perform well even in cases of limited training data. Such a breakthrough would greatly extend their applicability. However, we find that SAEs applied to spiking neural data are prone to a particular form of overfitting that cannot be detected using standard validation metrics, which prevents standard HP searches. We develop and test two potential solutions: an alternate validation method (“sample validation”) and a novel regularization method (“coordinated dropout”). These innovations prevent overfitting quite effectively, and allow us to test whether SAEs can achieve good performance on limited data through large-scale HP optimization. When applied to data from motor cortex recorded while monkeys made reaches in various directions, large-scale HP optimization allowed SAEs to better maintain performance for small dataset sizes. Our results should greatly extend the applicability of SAEs in extracting latent dynamics from sparse, multidimensional data, such as neural population spiking activity. 

   more » « less
3. Attractor-Like Dynamics Extracted from Human Electrocorticographic Recordings Underlie Computational Principles of Auditory Bistable Perception

   https://doi.org/10.1523/JNEUROSCI.1531-22.2023  

   Melland, Pake ; Curtu, Rodica ( March 2023 , The Journal of Neuroscience)

   In bistable perception, observers experience alternations between two interpretations of an unchanging stimulus. Neurophysiological studies of bistable perception typically partition neural measurements into stimulus-based epochs and assess neuronal differences between epochs based on subjects' perceptual reports. Computational studies replicate statistical properties of percept durations with modeling principles like competitive attractors or Bayesian inference. However, bridging neuro-behavioral findings with modeling theory requires the analysis of single-trial dynamic data. Here, we propose an algorithm for extracting nonstationary timeseries features from single-trial electrocorticography (ECoG) data. We applied the proposed algorithm to 5-min ECoG recordings from human primary auditory cortex obtained during perceptual alternations in an auditory triplet streaming task (six subjects: four male, two female). We report two ensembles of emergent neuronal features in all trial blocks. One ensemble consists of periodic functions that encode a stereotypical response to the stimulus. The other comprises more transient features and encodes dynamics associated with bistable perception at multiple time scales: minutes (within-trial alternations), seconds (duration of individual percepts), and milliseconds (switches between percepts). Within the second ensemble, we identified a slowly drifting rhythm that correlates with the perceptual states and several oscillators with phase shifts near perceptual switches. Projections of single-trial ECoG data onto these features establish low-dimensional attractor-like geometric structures invariant across subjects and stimulus types. These findings provide supporting neural evidence for computational models with oscillatory-driven attractor-based principles. The feature extraction techniques described here generalize across recording modality and are appropriate when hypothesized low-dimensional dynamics characterize an underlying neural system.

   SIGNIFICANCE STATEMENTIrrespective of the sensory modality, neurophysiological studies of multistable perception have typically investigated events time-locked to the perceptual switching rather than the time course of the perceptual states per se. Here, we propose an algorithm that extracts neuronal features of bistable auditory perception from largescale single-trial data while remaining agnostic to the subject's perceptual reports. The algorithm captures the dynamics of perception at multiple timescales, minutes (within-trial alternations), seconds (durations of individual percepts), and milliseconds (timing of switches), and distinguishes attributes of neural encoding of the stimulus from those encoding the perceptual states. Finally, our analysis identifies a set of latent variables that exhibit alternating dynamics along a low-dimensional manifold, similar to trajectories in attractor-based models for perceptual bistability.

    

   more » « less
4. A Generative Reinforcement Learning Framework for Predictive Analytics

   Skordilis, Erotokritos ; Moghaddass, Ramin. ; Farhat, Tazin. ( January 2023 , Proceedings annual Reliability and Maintainability Symposium)

   In this work, we present a new approach for latent system dynamics and remaining useful life (RUL) estimation of complex degrading systems using generative modeling and reinforcement learning. The main contributions of the proposed method are two-fold. First, we show how a deep generative model can approximate the functionality of high-fidelity simulators and, thus, is able to substitute expensive and complex physics-based models with data-driven surrogate ones. In other words, we can use the generative model in lieu of the actual system as a surrogate model of the system. Furthermore, we show how to use such surrogate models for predictive analytics. Our method follows two main steps. First, we use a deep variational autoencoder (VAE) to learn the distribution over the latent state-space that characterizes the dynamics of the system under monitoring. After model training, the probabilistic VAE decoder becomes the surrogate system model. Then, we develop a scalable reinforcement learning framework using the decoder as the environment, to train an agent for identifying adequate approximate values of the latent dynamics, as well as the RUL.To our knowledge, the method presented in this paper is the first in industrial prognostics that utilizes generative models and reinforcement learning in that capacity. While the process requires extensive data preprocessing and environment tailored design, which is not always possible, it demonstrates the ability of generative models working in conjunction with reinforcement learning to provide proper value estimations for system dynamics and their RUL. To validate the quality of the proposed method, we conducted numerical experiments using the train_FD002 dataset provided by the NASA CMAPSS data repository. Different subsets were used to train the VAE and the RL agent, and a leftover set was then used for model validation. The results shown prove the merit of our method and will further assist us in developing a data-driven RL environment that incorporates more complex latent dynamic layers, such as normal/faulty operating conditions and hazard processes. 

   more » « less
5. Deep echo state networks with uncertainty quantification for spatio‐temporal forecasting

   https://doi.org/10.1002/env.2553  

   McDermott, Patrick L. ; Wikle, Christopher K. ( December 2018 , Environmetrics)

   Long‐lead forecasting for spatio‐temporal systems can entail complex nonlinear dynamics that are difficult to specify a priori. Current statistical methodologies for modeling these processes are often highly parameterized and, thus, challenging to implement from a computational perspective. One potential parsimonious solution to this problem is a method from the dynamical systems and engineering literature referred to as an echo state network (ESN). ESN models usereservoir computingto efficiently compute recurrent neural network forecasts. Moreover, multilevel (deep) hierarchical models have recently been shown to be successful at predicting high‐dimensional complex nonlinear processes, particularly those with multiple spatial and temporal scales of variability (such as those we often find in spatio‐temporal environmental data). Here, we introduce a deep ensemble ESN (D‐EESN) model. Despite the incorporation of a deep structure, the presented model is computationally efficient. We present two versions of this model for spatio‐temporal processes that produce forecasts and associated measures of uncertainty. The first approach utilizes a bootstrap ensemble framework, and the second is developed within a hierarchical Bayesian framework (BD‐EESN). This more general hierarchical Bayesian framework naturally accommodates non‐Gaussian data types and multiple levels of uncertainties. The methodology is first applied to a data set simulated from a novel non‐Gaussian multiscale Lorenz‐96 dynamical system simulation model and, then, to a long‐lead United States (U.S.) soil moisture forecasting application. Across both applications, the proposed methodology improves upon existing methods in terms of both forecast accuracy and quantifying uncertainty.

    

   more » « less