Code360 powered by Coding Ninjas X Naukri.com. Code360 powered by Coding Ninjas X Naukri.com
Table of contents
1.
Retail Domain
1.1.
Target Time Series Dataset Type
1.2.
Related Time Series Dataset Type
1.3.
Item Metadata Dataset Type
2.
CUSTOM Domain
2.1.
Target Time Series Dataset Type
2.2.
Related Time Series Dataset Type
2.3.
Item Metadata Dataset Type
3.
INVENTORY_PLANNING Domain
3.1.
Target Time Series Dataset Type
3.2.
Related Time Series Dataset Type
3.3.
Item Metadata Dataset Type
4.
EC2 CAPACITY Domain
4.1.
Target Time Series Dataset Type
4.2.
Related Time Series Dataset Type
5.
WORK_FORCE Domain
5.1.
Target Time Series Dataset Type
5.2.
Related Time Series Dataset Type
5.3.
Item Metadata Dataset Type
6.
WEB_TRAFFIC Domain
6.1.
Target Time Series Dataset Type
6.2.
Related Time Series Dataset Type
6.3.
Item Metadata Dataset Type
7.
METRICS Domain
7.1.
Target Time Series Dataset Type
7.2.
Related Time Series Dataset Type
7.3.
Item Metadata Dataset Type
8.
Frequently Asked Questions 
8.1.
What method of forecasting Does Amazon use?
8.2.
How does Amazon forecast its sales?
8.3.
What is the P90 forecast on Amazon?
8.4.
Which cases apply to forecasting?
8.5.
What are three forecasting techniques?
9.
Conclusion
Last Updated: Mar 27, 2024

Amazon Forecast part 2

Author soham Medewar
0 upvote
Master Python: Predicting weather forecasts
Speaker
Ashwin Goyal
Product Manager @

This blog will be the continuation of amazon forecast part 1

Retail Domain

The following dataset types are supported by the RETAIL domain. We list essential and optional fields for each dataset type.

Target Time Series Dataset Type

The historical time series data for each item or product sold by the retail company is the target time series. The following fields must be filled out:

  • item_id (string) – A unique identification for the item or product for which you wish to forecast demand.
  • timestamp (timestamp)
  • demand (float) - The number of sales for that item at the timestamp (float). Amazon Forecast also generates a forecast for this target field.

The following dimension is optional and can be used to adjust the granularity of forecasting:

  • location (string) – The store where the item was purchased. Only use this if you have several stores/locations.

Only these required fields and optional dimensions should ideally be supplied.

Related Time Series Dataset Type

You can offer Amazon Forecast with associated time-series datasets, such as the item's price or the number of site hits on a specific date. The more details you supply, the more accurate the forecast will be. The following fields must be filled out:

  • item_id (string)
  • timestamp (timestamp)

The following items are optional but may help improve forecast results:

  • price (float) – The item's price at the time the timestamp was set.
  • promotion_applied (integer; 1 = true, 0 = false) A flag indicating whether or not the item was the subject of a marketing promotion at the time of the timestamp.

Your training data can include additional fields in addition to the mandatory and suggested optional fields. Provide the fields in your dataset's schema to include other fields in the dataset.

Item Metadata Dataset Type

This dataset offers Amazon Forecast with metadata (attributes) for the items whose demand is forecasted. The following fields must be filled out:

  • item_id = (string)

The following items are optional but may help improve forecast results:

  • category (string)
  • brand (string)
  • color (string)
  • genre (string)

Your training data can include additional fields in addition to the mandatory and suggested optional fields. To include more fields in the dataset, include the fields in a schema when creating the dataset.

Must read topic: Amazon Hirepro

CUSTOM Domain

Target Time Series Dataset Type

The following fields must be filled out:

  • item_id (string)
  • timestamp (timestamp)
  • target_value (floating-point integer) 

Related Time Series Dataset Type

The following fields must be filled out:

  • item_id (string)
  • timestamp (timestamp)

Item Metadata Dataset Type

Required Field:

  • item_id (string)

Optional Field:

  • category (string)
Get the tech career you deserve, faster!
Connect with our expert counsellors to understand how to hack your way to success
User rating 4.7/5
1:1 doubt support
95% placement record
Akash Pal
Senior Software Engineer
326% Hike After Job Bootcamp
Himanshu Gusain
Programmer Analyst
32 LPA After Job Bootcamp
After Job
Bootcamp

INVENTORY_PLANNING Domain

Use the INVENTORY_PLANNING domain to forecast raw material demand and determine how much inventory of a specific item to stock. It supports the dataset types listed below. We list essential and optional fields for each dataset type.

Target Time Series Dataset Type

The following fields must be filled out:

  • item_id (string)
  • timestamp (timestamp)
  • demand (float) - Amazon Forecast provides a forecast for this target field.

The following dimension is optional and can be used to adjust the granularity of forecasting:

  • location (string) - The distribution facility in which the item is stocked. Only use this if you have several stores/locations.

Related Time Series Dataset Type

The following fields must be filled out:

  • item_id (string)
  • timestamp (timestamp)

The following items are optional but may help improve forecast results:

  • price (float) - The price of the item

Item Metadata Dataset Type

The following fields must be filled out:

  • item_id (string)

The following items are optional but may help improve forecast results:

  • category (string) - The item's category.
  • brand (string) - The item's brand.
  • lead_time (string) - The number of days required to manufacture the item.
  • order_cycle (string) - The order cycle begins with the commencement of work and ends when the item is ready for delivery.
  • safety_stock (string) - The bare minimum of inventory to have on hand for that item.

EC2 CAPACITY Domain

Amazon Forecast EC2 capacity using the EC2 CAPACITY domain. It supports the dataset types listed below. We list essential and optional fields for each dataset type.

Target Time Series Dataset Type

The following fields must be filled out:

  • instance_type (string) – The instance's type (for example, c5.xlarge).
  • timestamp (timestamp)
  • number_of_instances (integer) – The number of instances of that specific instance type consumed at that moment. Amazon Forecast provides a forecast for this target field.

The following dimension is optional and can be used to adjust the granularity of forecasting:

  • location (string) – An AWS Region, such as us-west-2 or us-east-1, can be provided. Only use this if you're modelling numerous Regions.

Related Time Series Dataset Type

The following fields must be filled out:

  • instance_type (string)
  • timestamp (timestamp)

WORK_FORCE Domain

Use the WORK_FORCE domain to forecast workforce demand. It supports the dataset types listed below. We list essential and optional fields for each dataset type.

Target Time Series Dataset Type

The following fields must be filled out:

  • workforce_type (string) - The type of labour to be forecasted. For instance, contact center or fulfillment center labour need.
  • timestamp (timestamp)
  • workforce_demand (floating-point integer) - The target field for which Amazon Forecast produces a forecast.

The following dimension is optional and can be used to adjust the granularity of forecasting:

  • location (string) - The location in which labour resources are searched. If you have many stores/locations, this should be used.

Related Time Series Dataset Type

The following fields must be filled out:

  • workforce_type (string)
  • timestamp (timestamp)

Item Metadata Dataset Type

The following fields must be filled out:

  • workforce_type (string)

The following items are optional but may help improve forecast results:

  • wages (float) – The average pay for that worker type.
  • shift_length (string) – The shift's length.
  • location (string) - The workforce's location.

WEB_TRAFFIC Domain

To forecast web traffic to a website or set of web properties, use the WEB TRAFFIC domain. It supports the dataset types listed below.

Target Time Series Dataset Type

The following fields must be filled out:

  • item_id (string) — A unique identifier for each forecasted web property.
  • timestamp (timestamp) 
  • value (float) – This is the target field for which Amazon Forecast makes a forecast.

Related Time Series Dataset Type

The following fields must be filled out:

  • item_id (string)
  • timestamp (timestamp)

Item Metadata Dataset Type

The following fields must be filled out:

  • item_id (string)

The following field is optional but may help improve forecast results:

  • classification (string)

METRICS Domain

Forecast indicators like revenue, sales, and cash flow using the METRICS domain. It supports the dataset types listed below.

Target Time Series Dataset Type

The following fields must be filled out:

  • metric_name (string)
  • timestamp (timestamp)
  • metric_value (floating-point integer) - The target field for which Amazon Forecast makes a forecast (for example, the amount of revenue generated on a particular day).

Related Time Series Dataset Type

The following fields are required:

  • metric_name (string)
  • timestamp (timestamp)

Item Metadata Dataset Type

The following fields must be filled out:

  • metric_name (string)

The following items are optional but may help improve forecast results:

  • category (string)

Frequently Asked Questions 

What method of forecasting Does Amazon use?

Amazon Forecast is a time-series forecasting service based on machine learning (ML) and built for business metrics analysis.

How does Amazon forecast its sales?

Amazon Forecast forecasts future business outcomes for FBA sellers, such as resource requirements, product demand, and financial success, using machine learning technologies. This fully-managed service's built-in technologies analyze your Amazon FBA historical data to find patterns and forecast future sales.

What is the P90 forecast on Amazon?

There's a 90% probability that weekly consumer demand will be at the level indicated and a 10% chance Amazon will buy more.

Which cases apply to forecasting?

Time-series forecasting has various applications in planning. The most popular application is to compare Predictive Planning's statistical projections to your own forecast.

What are three forecasting techniques?

Qualitative approaches, time series analysis and projection, and causal models are the three main types.

Conclusion

In this article, we have discussed the following topics:

  • Features of Amazon Forecast
  • Working of Amazon Forecast
  • Predefined Dataset Domains and Dataset Types
  • Using item metadata datasets

After reading about the attributes of a file, are you not feeling excited to read/explore more articles on the topic of AWS? Don't worry; Coding Ninjas has you covered. To learn, see AWS featuresGlobal Infrastructure concept in AWS, and AWS SAML.

Check out the Amazon Interview Experience to learn about Amazon’s hiring process.

You can also consider our Online Coding Courses such as the Machine Learning Course to give your career an edge over others.

Do upvote our blogs if you find them helpful and engaging!

Happy Coding!

Next article
Amazon Comprehend
Live masterclass