Entries for Net Load Forecasting Prize

We plan to use a data-driven approach using machine/deep learning techniques to forecast net loads.

We use a two-tier ML modeling approach: project the available resources and forecast electrical demand at a point on the grid line.

Use dynamic, time-series learning to accurately predict net load

A machine learning pipeline to predict net load

Statisical and data analysis using neural pattern matching synthesis

Using GBM and deep learning models

Use vector autoregressive models and machine learning to generate hundreds of future netload scenarios; then take the quantiles from these.

Use our expertise to develop probabilistic models, streamline forecasting processes, and drive the transition to clean energy.

We have put together a diverse team to perform day-ahead net load forecasting from historical data using deep sequential learning

The forecasting algorithm is based on the latest machine-learning tools for time-series forecasts augmented with weather-based predictors.

Machine learning techniques combined with statistical methods will be used to be a probabilistic net load prediction model.

Locbit allows companies to optimize their energy usage. We plan to expand that to an energy load forecasting system commercially.

We develop machine learning models for probabilistic forecasting.

Decomposing net load into three parts, predicting them separately and finally using copula theory to generate probabilistic forecasting.

Implementation of a range of machine & deep learning models with key feature engineering & data processing to forecast net load percentiles.