Feature List
  • 8.x
    • Team Feature
      • Data-Prep Features
      • Pipeline Features
        • What's New Window Pop-up
        • Failure Db-Sync
        • Testing for Kafka 3.1.0
        • Python 3.10 R&D
        • Job History
        • SPC (Single Page for Configuration)
        • Python Jobs
        • Failure Alerts
        • Event Channel Alerts
        • Pipeline Error handling
        • Pipeline : PySpark Component (PySpark Job)
      • Dashboard Charting
        • Widget as component
        • Knowledge Graph Chart
          • Sample Library based code
        • Word Cloud
        • Tile component
        • Sankey Chart
        • Model as API Connector
        • Dataprep recipe in Dataset selection
        • Decomposition Enhancment
      • Python Upgrade
        • Core Platform : Data Services
        • Core Platform : Data Catalog
        • Core Platform : Data Center
        • Data Science Lab
      • Sonar Code Scan automation by DevOps
      • DS Lab PySpark Project.
      • Core Platform
        • Tag Feature For Data Connector , Dataset , DataStore etc..
        • DataStore & Metadatastore Migration
        • MongoDB & ClickHouse Support For DataSheet
        • Data As API WorkBench
        • Pagination in Home , DataCenter , Dataset , DataStore etc..
        • Sharing Data Connector & Data Set with View or Edit Permission.
        • Core Monitoring & Alerting
      • Data Science Lab
        • Auto Forecasting Requirements
          • User Input
          • Forecasting Method
          • Explainability
        • DSLAB Sprint May1-2023-May12-2023
        • DS LAB Sprint Apr10-Apr21
        • Provide Static Variables for DSLAB Component In AutoML
        • Scheduler For DSLAB Scripts
        • Optimisation of Model Explainability code
    • QA
    • DevOps
Powered by GitBook
On this page
  1. 8.x
  2. Team Feature
  3. Data Science Lab
  4. Auto Forecasting Requirements

User Input

Input required form Users:

  • Target column

  • Date Column

  • Forecast Horizon

  • Static features : The time-independent attributes (metadata) of a time series. These may include information such as:

    • location, where the time series was recorded (country, state, city)

    • fixed properties of a product (brand name, color, size, weight)

    • store ID or product ID

  • Known Covariates : are known for the entire forecast horizon, such as

    • holidays

    • day of the week, month, year

    • promotions

  • Past Covariates : are only known up to the start of the forecast horizon, such as

    • sales of other products

    • temperature, precipitation

    • transformed target time series

  • Handling Missing Data :

    • Completely ignore the time index

    • Extend the index and fill missing values

PreviousAuto Forecasting RequirementsNextForecasting Method

Last updated 1 year ago