Temporal Multimodal Multivariate Learning
Hyoshin Park,Justice Eric Darko,Niharika M. Deshpande,Venktesh Pandey,H. Su,Masahiro Ono,Dedrick Barkely,Larkin Folsom,Derek J. Posselt,Steve Chien +9 more
- 14 Jun 2022
TL;DR:
read more
Abstract: We introduce temporal multimodal multivariate learning, a new family of decision making models that can indirectly learn and transfer online information from simultaneous observations of a probability distribution with more than one peak or more than one outcome variable from one time stage to another. We approximate the posterior by sequentially removing additional uncertainties across different variables and time, based on data-physics driven correlation, to address a broader class of challenging time-dependent decision-making problems under uncertainty. Extensive experiments on real-world datasets ( i.e., urban traffic data and hurricane ensemble forecasting data) demonstrate the superior performance of the proposed targeted decision-making over the state-of-the-art baseline prediction methods across various settings.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Physics-Informed Deep Learning with Kalman Filter Mixture for Traffic State Prediction
Niharika M. Deshpande,Hyoshin Park +1 more
TL;DR: This study proposes a novel physics-informed deep learning model, PI-GRNN, integrated with Kalman Filter to predict traffic states, capturing epistemic uncertainty through dynamic spatiotemporal correlations and periodic corrections, outperforming benchmark approaches with real-world traffic data.
4
Advancing Temporal Multimodal Learning with Physics Informed Regularization
22 Mar 2023
TL;DR: In this paper , a physics-informed and regularized prediction model is developed that shares observations across similarly distributed network segments across time and space to estimate multimodal distributions of travel times from real-world data.
2
Advancing Temporal Multimodal Learning with Physics Informed Regularization
Niharika M. Deshpande,Hyoshin Park,Venktesh Pandey,Gyugeun Yoon +3 more
- 22 Mar 2023
TL;DR: In this paper , a physics-informed and regularized prediction model is developed that shares observations across similarly distributed network segments across time and space to estimate multimodal distributions of travel times from real-world data.
1
A New Predictive Maintenance Approach: novel integration of GMM-LSTM for prediction of latent state and failure location of Rotating Machinery
Arman Ghavidel,Hyoshin Park,Samuel Kovacic,Andres Sousa‐Poza +3 more
- 23 Jun 2024
TL;DR: This paper introduces a novel predictive maintenance approach integrating Gaussian Mixture Models (GMMs) and Long Short-Term Memory (LSTM) networks for robust feature extraction and precise failure location prediction in rotating machinery, validated on distinct datasets with high precision and recall rates.
Physics-Informed Deep Learning with Kalman Filter Mixture: A New State Prediction Model
Niharika M. Deshpande,Hyoshin Park +1 more
- 13 Mar 2024
TL;DR: This study captures unobserved heterogeneity in travel time by considering distinct peaks in the probability density function, which it refers to as multimodal probability distribution while establishing causation through physics-based principles to construct a dynamic graph of a traffic network.
References
Time-dependent routing problems
TL;DR: A comprehensive review of travel time modelling, applications and solution methods is presented and a first classification in point-to-point and multiple-point problems is made with respect to the quality and evolution of information.
255
Information and Sufficiency
H. Heyer
- 01 Jan 1982
TL;DR: This section introduces the decision theoretic basis to the approach to mathematical statistics and contains the fundamental definitions, some interpretations, and some theorems of an introductory nature designed to give a first insight into the theory of comparison of experiments.
176
•Posted Content
Horizon: Facebook’s Open Source Applied Reinforcement Learning Platform
Jason Gauci,Edoardo Conti,Yitao Liang,Kittipat Virochsiri,Yuchen He,Zachary Kaden,Vivek Narayanan,Xiaohui Ye +7 more
TL;DR: Facebook's open source applied reinforcement learning (RL) platform Horizon is presented, which contains workflows to train popular deep RL algorithms and includes data preprocessing, feature transformation, distributed training, counterfactual policy evaluation, optimized serving, and a model-based data understanding tool.
Multistate Model for Travel Time Reliability
TL;DR: A novel multistate model is proposed for travel time modeling and reporting that provides much improved model fitting as compared with single-mode models and provides a connection between travel time distributions and the underlying travel time state.
120
Formulation, existence, and computation of boundedly rational dynamic user equilibrium with fixed or endogenous user tolerance
TL;DR: In this paper, a unified approach for modeling both BR-DUE and variable tolerance dynamic user equilibrium is presented, which makes significant contributions to the model formulation, analysis of existence, solution characterization, and numerical computation of such problems.
100