microprediction

microprediction / LearningKalman / 0.1.2

README.md

Overview

Stateful Kalman filter

Applicable Scenarios and Problems

Denoising lenghty univariate time series

Usage

Input

Describe the input fields for your algorithm. For example:

ParameterDescription
idUnique series identifier
valueNoisy observation

Output

ParameterDescription
valuePrediction of next point
PVariance of next point
QTime step variance estimate
RNoise variance estimate

Parameters Q and R are notoriously commingled so don't get your hopes up