Data Denoising with Weighted Forward Moving AverageΒΆ

Non-Causal Step

This step denoises data similarly to the pre-processing step based on the Forward Moving Average algorithm, but additionally gives a weighting factor to each value within the filter width. Note that the applied algorithm doesn’t need the timestamp information for the denoising procedure.

Input Parameters

  1. One-dimensional time series data
  2. Filter width
  3. Weights for values in the filter width

Output Parameters

  1. Denoised time series data

Workflow

../../../../_images/workflow16.svg

Algorithm

Weighted Forward Moving Average

References