In this case study, we will be using the Electricity Load dataset from UCI Machine Learning Repository. The dataset describes the 15-minutes load for 370 different clients in kW. Each column represents one client. The dataset has electricity consumption for these 370 houses every 15 mins starting from 2011-01-01, till 2015-01-01. Some clients are created after 2011. In these cases, consumption was considered zero.
We need to build a forecasting model to forecast the electricity load of different clients (i.e., forecasting multiple time series at once) for a certain period in the future.
This algorithm is known as Multivariate Singular Spectrum Analysis (mSSA). This is a matrix factorization based method that can help in -
- Time-series forecasting and
- Time series imputation
The python library to use this algorithm is not currently present in PyPi. The code repository is present in this GitHub repository: https://github.com/AbdullahO/mSSA. So we will be using a different way to install the library on your local machine. Below we have provided the steps to follow to install this library.