Select your cookie preferences

We use essential cookies and similar tools that are necessary to provide our site and services. We use performance cookies to collect anonymous statistics, so we can understand how customers use our site and make improvements. Essential cookies cannot be deactivated, but you can choose “Customize” or “Decline” to decline performance cookies.

If you agree, AWS and approved third parties will also use cookies to provide useful site features, remember your preferences, and display relevant content, including relevant advertising. To accept or decline all non-essential cookies, choose “Accept” or “Decline.” To make more detailed choices, choose “Customize.”
Sign in
Sign in
or
Create a new account
About
Categories
Delivery Methods
Solutions
Resources
 Your Saved List Become a Channel Partner Sell in AWS Marketplace Amazon Web Services Home Help

Amazon Sagemaker

Amazon SageMaker is a fully-managed platform that enables developers and data scientists to quickly and easily build, train, and deploy machine learning models at any scale. With Amazon SageMaker, all the barriers and complexity that typically slow down developers who want to use machine learning are removed. The service includes models that can be used together or independently to build, train, and deploy your machine learning models.

product logo

Yule-Walker-PCA Autoregression (YWpcAR) Free trial

By:
i4cast LLC
Latest Version:
0.3.0
Yule-Walker-PCA Autoregressive Model (YWpcAR) to analyze and forecast many time-series individually with evolution of hidden components.
Continue to Subscribe

    Product Overview

    The Yule-Walker-PCA Autoregressive Model (YWpcAR) algorithm is developed to simultaneously analyze and forecast many time-series individually, assuming each time-series is influenced by evolution of "hidden components" (resulted from PCA). Here PCA standards for "principal components analysis". Different time-series is influenced by different sets of hidden components (PCs). By applying objective data-driven constraints, the YWpcAR algorithm can estimate the influences of longer histories of the PCs. The algorithm accommodates wider ranges of values of model learning parameters. The wider ranges can further enhance the power of machine learning. Current version of the YWpcAR algorithm estimates: (a) autoregressive coefficients of time-series, (b) filter coefficients to generate unobserved component (sum of PCs), (c) time-series of the unobserved component, and (d) forecasts of the observed time-series. Other estimates will be added in the future releases.

    Key Data

    Version
    0.3.0
    Usage Instructions
    Release Notes
    Show other versions
    By
    i4cast LLC
    Categories
    Time-series Forecasting
    Dimensionality Reduction
    Type
    Algorithm
    Fulfillment Methods
    Amazon SageMaker

    Highlights

    • Introducing PCA into YW-AR modeling:

      1. Applying principal components analysis (PCA) to sample variance-autocovariance matrix, C, in Yule-Walker (YW) equation of autoregressive (AR) model.
      2. Replacing elements of the matrix C by PCA-based common components.
      3. Replacing elements of the matrix and vector in the YW equation by the PCA-based common components of C.
      4. Estimating AR model coefficients by the PCA-based YW equation.
      5. In time-series forecast with the YW-PCA AR (YWpcAR) model, replacing observed time-series data by unobserved components associated with the PCs.

    • Benefits of introducing PCA into YW-AR modeling:

      1. Noise reduction due to dimension reduction when the number of PCs, m, smaller than the autoregressive order, p.
      2. Avoiding over-fitting when estimating long-memory AR model of relatively larger value of order p.

    Not quite sure what you’re looking for? AWS Marketplace can help you find the right solution for your use case. Contact us

    Pricing Information

    Use this tool to estimate the software and infrastructure costs based your configuration choices. Your usage and costs might be different from this estimate. They will be reflected on your monthly AWS billing reports.

    Contact us to request contract pricing for this product.


    Estimating your costs

    Choose your region and launch option to see the pricing details. Then, modify the estimated price by choosing different instance types.

    Version
    Region

    Software Pricing

    Algorithm Training$0.10/hr

    running on ml.m5.xlarge

    Model Realtime Inference$0.10/hr

    running on ml.m5.xlarge

    Model Batch Transform$0.10/hr

    running on ml.m5.xlarge

    Infrastructure Pricing

    With Amazon SageMaker, you pay only for what you use. Training and inference is billed by the second, with no minimum fees and no upfront commitments. Pricing within Amazon SageMaker is broken down by on-demand ML instances, ML storage, and fees for data processing in notebooks and inference instances.
    Learn more about SageMaker pricing

    SageMaker Algorithm Training$0.23/host/hr

    running on ml.m5.xlarge

    SageMaker Realtime Inference$0.23/host/hr

    running on ml.m5.xlarge

    SageMaker Batch Transform$0.23/host/hr

    running on ml.m5.xlarge

    About Free trial

    Try this product for 100 days. There will be no software charges, but AWS infrastructure charges still apply. Free Trials will automatically convert to a paid subscription upon expiration.

    Algorithm Training

    For algorithm training in Amazon SageMaker, the software is priced based on hourly pricing that can vary by instance type. Additional infrastructure cost, taxes or fees may apply.
    InstanceType
    Algorithm/hr
    ml.m4.4xlarge
    $0.10
    ml.c5n.18xlarge
    $0.10
    ml.g4dn.4xlarge
    $0.10
    ml.m5.4xlarge
    $0.10
    ml.m4.16xlarge
    $0.10
    ml.m5.2xlarge
    $0.10
    ml.p3.16xlarge
    $0.10
    ml.g5.xlarge
    $0.10
    ml.g5.12xlarge
    $0.10
    ml.g4dn.2xlarge
    $0.10
    ml.g5.4xlarge
    $0.10
    ml.m4.2xlarge
    $0.10
    ml.c5.2xlarge
    $0.10
    ml.c4.2xlarge
    $0.10
    ml.g4dn.12xlarge
    $0.10
    ml.p4d.24xlarge
    $0.10
    ml.m4.10xlarge
    $0.10
    ml.m5.24xlarge
    $0.10
    ml.g4dn.xlarge
    $0.10
    ml.g5.48xlarge
    $0.10
    ml.g4dn.16xlarge
    $0.10
    ml.m5.12xlarge
    $0.10
    ml.p3dn.24xlarge
    $0.10
    ml.p2.16xlarge
    $0.10
    ml.c4.4xlarge
    $0.10
    ml.g5.8xlarge
    $0.10
    ml.m5.xlarge
    Vendor Recommended
    $0.10
    ml.c5.9xlarge
    $0.10
    ml.g5.16xlarge
    $0.10
    ml.m4.xlarge
    $0.10
    ml.c5.4xlarge
    $0.10
    ml.p3.8xlarge
    $0.10
    ml.c4.8xlarge
    $0.10
    ml.g4dn.8xlarge
    $0.10
    ml.p2.8xlarge
    $0.10
    ml.c5n.2xlarge
    $0.10
    ml.c5n.9xlarge
    $0.10
    ml.c5.18xlarge
    $0.10
    ml.g5.2xlarge
    $0.10
    ml.c5n.4xlarge
    $0.10
    ml.g5.24xlarge
    $0.10

    Usage Information

    Training

    The Yule-Walker-PCA Autoregressive Model (YWpcAR) algorithm is developed to simultaneously analyze and forecast many time-series individually, assuming each time-series is influenced by evolution of "hidden components" (resulted from PCA).

    The YWpcAR algorithm takes, as input data, multiple time-series data contained in a CSV (comma separated value) data table, in a format of a CSV text-string or a CSV text-file. Each row of the data table is for values of an individual time-series (TS). Each column is for values of all time-series at a specific moment in time.

    Metrics

    Name
    Regex
    projection_coefficient_#_13
    projection_coefficient_#_13=(.*?);
    projection_coefficient_#_26
    projection_coefficient_#_26=(.*?);
    projection_coefficient_#_52
    projection_coefficient_#_52=(.*?);

    Channel specification

    Fields marked with * are required

    train

    *
    Training dataset
    Input modes: File
    Content types: text/csv
    Compression types: None

    Hyperparameters

    Fields marked with * are required

    len_learn_window

    *
    Length of moving/rolling time window for model learning/estimate
    Type: Integer
    Tunable: Yes

    ar_order

    *
    Order, p, of AR (autoregressive) model
    Type: Integer
    Tunable: Yes

    num_pcs

    *
    Number of principal components (PCs) for YW-PCA AR model
    Type: Integer
    Tunable: Yes

    max_forecast_step

    *
    Maximum number of future/forward/lead steps to be forecasted
    Type: Integer
    Tunable: No

    target_type

    *
    Type of time-series value target to be forecasted
    Type: Categorical
    Tunable: No

    fwd_cumsum

    *
    Forward cumulative summation in forecasts (True/False)
    Type: Categorical
    Tunable: No

    num_forecasts

    *
    Number of rolling forecasts for model evaluation
    Type: Integer
    Tunable: No

    half_life_list

    *
    List of half-life of time-weightings applying to time-series forecast model evaluation
    Type: FreeText
    Tunable: No

    eval_metric_list

    *
    List of model evaluation metrics
    Type: FreeText
    Tunable: No

    alt_ar_order

    AR orders for some specific time-series: defined by paires of key (label of time-series) and value (AR order)
    Type: FreeText
    Tunable: No

    alt_num_pcs

    Numbers of PCs for some specific time-series: defined by paires of key (label of time-series) and value (number of PCs)
    Type: FreeText
    Tunable: No

    Model input and output details

    Input

    Summary

    The YWpcAR (Yule-Walker-PCA Autoregressive model) algorithm takes, as input data, multiple time-series data contained in a CSV (comma separated value) data table, in a format of a CSV text-string or a CSV text-file.

    Each row of the data table is for values of an individual time-series (TS). Row header is the label or symbol of the time-series. Each column is for values of all time-series at a specific moment in time. Column header is the time-index or time-stamp of the moment.

    Input MIME type
    text/csv
    Sample input data
    Expand all input descriptions
    Field name
    Values of time stamp
    Description
    Each row of the data table is for values of an individual time-series (TS). Row header is the label or symbol of the time-series. Each column is for values of all time-series at a specific moment in time. Column header is the time-index or time-stamp of the moment. The first data column is for the earliest time and the last column for the most recent time. The current version of YWpcAR requires equally spaced time-stamps.
    Required
    Yes
    Data type
    FreeText

    Output

    Summary

    YWpcAR model output are multi-step or multi-horizon forecasts of all time-series. Additional outputs include a variety of parameter/coefficient matrixes estimated by the PCA-based autoregressive (AR) model.

    Output in format of CSV tables can be used to make quick review by using a spreadsheet application. Output in format of JSON strings can be used as input data for further analysis.

    Output MIME type
    text/csv, application/json
    Sample output data
    Expand all output descriptions
    Field name
    forecast
    Description
    YWpcAR model output are multi-step or multi-horizon forecasts of all time-series. Additional outputs include a variety of parameter/coefficient matrixes estimated by the PCA-based autoregressive (AR) model. Output in format of CSV tables can be used to make quick review by using a spreadsheet application. Output in format of JSON strings can be used as input data for further analysis.
    Always returned
    Yes
    Data type
    FreeText

    Sample notebook

    Additional Resources

    End User License Agreement

    By subscribing to this product you agree to terms and conditions outlined in the product End user License Agreement (EULA)

    Support Information

    Yule-Walker-PCA Autoregression (YWpcAR)

    For questions or call-back number, please send email to i4cast LLC at prod.i4cast@gmail.com.

    AWS Infrastructure

    AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.

    Learn More

    Refund Policy

    We offer full refund for academic works. Other refunds are offered according to common practices.

    Customer Reviews

    There are currently no reviews for this product.
    View all