Analysis and Forecast
 
 
 
 
 
Time series analysis and forecast
 
RSS channel
 
 
 
 
 
Forecast Time Series

 

Caterpillar SSA - time series analysis and forecast

 

Features of Caterpillar 3.40

Analysis of one-dimensional time series
  1. Decomposition of one-dimensional time series into eigentriples (eigenvalues, eigenvectors and principal components)
  2. Convenient graphical visualization of results for identification of the eigentriples corresponding to trend, periodicities, noise
  3. Grouping of eigentriples that leads to expansion of the time series into additive components
  4. Reconstruction of time series components (trend, oscillations, periodicities, noise) by choice of eigentriples
  5. Residual analysis
Time series analysis

 

Forecast of one-dimensional time series
  1. Approximation (local) of time series by finite-rank series
  2. Forecast by vector and recurrent methods
  3. Analyzing the linear recurrent formula used for the recurrent forecast method
  4. Confidence intervals by empirical and bootstrap methods
  5. Construction of envelopes (channels)
  6. Testing the forecast results on validation period
Time series forecast

 

Change-point detection for one-dimensional time series
  1. Change-point detection by comparing the 'Caterpillar-SSA' structures of the base and test time series intervals
  2. Construction of heterogeneity matrix and detection functions
  3. Analyzing the found structural changes by moving root and modulus functions
Change-point detection for time series

 

Multichannel Analysis/Forecast of time series
  1. Simultaneous decomposition of several one-dimensional time series into common eigentriples (eigenvalues, eigenvectors and principal components)
  2. Convenient graphical visualization of results for identification of the eigentriples corresponding to trend, common periodicities, noise
  3. Grouping of eigentriples that leads to expansion of the time series into additive components
  4. Reconstruction of the time series components (trend, oscillations, periodicities, noise) by choice of eigentriples
  5. Approximation (local) of time series by finite-rank series
  6. Forecast by vector and recurrent methods
  7. Testing the forecast results on validation period
Multichannel Analysis/Forecast of Time Series

 

 

Features of CatMV 1.0

Main features:
  1. Decomposition of time series with missing values
  2. Extraction of trend and periodic components
  3. Approximation of the extracted component by a time series of finite rank with the help of Cadzow iterations method
  4. Filling in the missing data in the extracted component
  5. Forecasting the extracted component by setting missing values after non-missing ones
Additional features:
  1. Navigation through the program stages/results
  2. Loading data from file
  3. Simulation of model input data
  4. Editing the loaded data
  5. Saving results to a text file (only for registered users)
  6. Comparison of the reconstructed time series with a specified time series
  7. Editing graphics options by double click

 

Features of CatSSA 2.0 (DLL)

  1. Decomposition of one-dimensional time series into eigentriples (eigenvalues, eigenvectors and principal components) by Basic or Toeplitz SSA
  2. Reconstruction of time series components (trend, oscillations, periodicities, noise) by choice of eigentriples
  3. Forecast by the recurrent and vector methods

 

Features of Caterpillar 1.00

  1. Decomposition of one-dimensional time series into eigentriples (eigenvalues, eigenvectors and principal components)
  2. Convenient graphical visualization of results for identification of the eigentriples corresponding to trend, periodicities, noise
  3. Reconstruction of time series components (trend, oscillations, periodicities, noise) by choice of eigentriples

 
 
 
Copyright © 1996-2015 GistaT Group. All rights reserved.
 
 
Analysis Time Series