Probabilistic Forecasting of Sensory Data with Generative Adversarial Networks - ForGAN

Abstract

Time series forecasting is one of the challenging problems for humankind. Traditional forecasting methods using mean regression models have severe shortcomings in reflecting real-world fluctuations. While new probabilistic methods rush to rescue, they fight with technical difficulties like quantile crossing or selecting a prior distribution. To meld the different strengths of these fields while avoiding their weaknesses as well as to push the boundary of the state-of-the-art, we introduce ForGAN - one step ahead probabilistic forecasting with generative adversarial networks. ForGAN utilizes the power of the conditional generative adversarial network to learn the data generating distribution and compute probabilistic forecasts from it. We argue how to evaluate ForGAN in opposition to regression methods. To investigate probabilistic forecasting of ForGAN, we create a new dataset and demonstrate our method abilities on it. This dataset will be made publicly available for comparison. Furthermore, we test ForGAN on two publicly available datasets, namely Mackey-Glass dataset and Internet traffic dataset (A5M) where the impressive performance of ForGAN demonstrate its high capability in forecasting future values.

References in corpus (27)

• Conditional Generative Adversarial Nets
• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning
• Expectation Propagation for approximate Bayesian inference
• InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
• Machine Learning for Integrating Data in Biology and Medicine: Principles, Practice, and Opportunities
• Adversarial Feature Learning
• Wasserstein GAN
• C-RNN-GAN: Continuous recurrent neural networks with adversarial training
• Wind Energy: Forecasting Challenges for Its Operational Management
• Deep Gaussian Processes
• Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs
• Conditional Generative Adversarial Networks for Speech Enhancement and Noise-Robust Speaker Verification
• Anomaly Detection with Generative Adversarial Networks for Multivariate Time Series
• PhaseLink: A Deep Learning Approach to Seismic Phase Association
• Adversarial Learning for Neural Dialogue Generation
• Optimal Data-Based Binning for Histograms
• Probabilistic projections of HIV prevalence using Bayesian melding
• Nonlinear Systems Identification Using Deep Dynamic Neural Networks
• Conditional Transformation Models
• Language Generation with Recurrent Generative Adversarial Networks without Pre-training
• ML for Flood Forecasting at Scale
• A Multi-Scheme Ensemble Using Coopetitive Soft-Gating With Application to Power Forecasting for Renewable Energy Generation
• Model-Free Renewable Scenario Generation Using Generative Adversarial Networks
• Predicting Inpatient Discharge Prioritization With Electronic Health Records
• CRED: A Deep Residual Network of Convolutional and Recurrent Units for Earthquake Signal Detection
• Accurate, Data-Efficient Learning from Noisy, Choice-Based Labels for Inherent Risk Scoring

Cited by in corpus (7)

• Deep Learning for Time Series Forecasting: A Survey
• A review of predictive uncertainty estimation with machine learning
• Generative Adversarial Networks for Spatio-temporal Data: A Survey
• Random Noise vs State-of-the-Art Probabilistic Forecasting Methods : A Case Study on CRPS-Sum Discrimination Ability
• A Bayesian Generative Adversarial Network (GAN) to Generate Synthetic Time-Series Data, Application in Combined Sewer Flow Prediction
• If You Like It, GAN It. Probabilistic Multivariate Times Series Forecast With GAN
• Adversarial Machine Learning in Text Analysis and Generation