Detrending daily natural gas consumption series to improve short-term forecasts
Estimating the Extreme Low-Temperature Event using Nonparametric Methods
Mathematical Models for Natural Gas Forecasting
The Impact of Additional Weather Inputs on Gas Load Forecasting
Forecasting Natural Gas Design Day Demand from Historical Monthly Data
Forecasting Natural Gas Demand: The Role of Physical and Economic Factors
A Survey of Different Prediction Models & Role of Artificial Neural Networks for Natural Gas Consumption
Study on the relationship between the energy consumption and economic system of Jiangsu Province base on grey relational analysis
Predictive temporal patterns detection in multivariate dynamic data system
Identification of Non-Temperature-Sensitive Natural Gas Customers and Forecasting Their Demand
Consistent Forecasting of Severe Winter Gas Demand
Combination of artificial neural-network forecasters for prediction of natural gas consumption
Disaggregating time series data for energy consumption by aggregate and individual customer
Improving Daily Natural Gas Forecasting by Tracking and Combining Models
Grey Predictive on Natural Gas Consumption and Production in China
Exploiting Domain Knowledge to Forecast Heating Oil Consumption
Using Evolutionary Programming to Increase the Accuracy of an Ensemble Model For Energy Forecasting
Estimation of Natural Gas Demand in Industry Sector of Iran: A Nonlinear Approach
Natural Gas Distribution System: Design Aspects
Trading by Numbers: Scoring Strategies for Every Market
The GasDay Project at Marquette University: A Learning Laboratory in a Functioning Business
Neural Network based Forecasting Model for Natural Gas Consumption
Deep Neural Network Regression for Short-Term Load Forecasting of Natural Gas
Identifying extreme cold events using phase space reconstruction
Detrending Daily Natural Gas Demand Data Using Domain Knowledge
Modeling the Tail Distribution and Ratemaking: An Application of Extreme Value Theory
Deep Neural Network Regression as a Component of a Forecast Ensemble
Transforming analogous time series data to improve natural gas demand forecast accuracy
Predicting rare events: Evaluating systemic and idiosyncratic risk ☆
Using Surrogate Data to Mitigate the Risks of Natural Gas Forecasting on Unusual Days
A Conceptual Study of Model Selection in Classification - Multiple Local Models vs One Global Model
Deep Neural Networks As Time Series Forecasters of Energy Demand
Modeling Extreme Climate Events: Two Case Studies in Mexico
Detection of outliers in time series data
Improving gas demand forecast during extreme cold events
Statistical models of reconstructed phase spaces for signal classification
A Study of the Severity of the Midwestern Winters of 1977 and 1978 Using Heating Degree Days Determined from Both Measured and Wind Chill Temperatures
Blending as a Time Series Forecasting Tool
Least-Squres Polynomial Smoothing for Inductively Coupled Plasma Atomic Emission Spectra
Data cleaning in the energy domain
Long-Range Forecasting for a Consumer Durable in an International Market
This paper presents a novel detrending algorithm that allows long-term natural gas demand signals to be used effectively to generate high quality short-term natural gas demand forecasting models. Short data sets in natural gas forecasting inadequately represent the range of consumption patterns necessary for accurate short-term forecasting. In contrast, longer data sets present a wide range of customer characteristics, but their long-term historical trends must be adjusted to resemble recent data before models can be developed. Our approach detrends historical natural gas data using domain knowledge. Forecasting models trained on data detrended using our algorithm are more accurate than models trained using nondetrended data or data detrended by benchmark methods. Forecasting accuracy improves using detrended longer-term signals, while forecast accuracy decreases using non-detrended long-term signals.