Data estimation of historical data that identifies fine grained time period data and coarse grained time period data and applies correlated slopes to time periods

• US 9,104,964 B1
• Filed: 11/16/2012
• Issued: 08/11/2015
• Est. Priority Date: 11/16/2012
• Status: Expired due to Fees

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Siddiqui, Muhammad Ali, Verma, Sanjeev Kewal, Crissman, Charles Graham Haver, Veronda, Mark Christopher
• Primary Examiner(s): Vincent, David
• Application Number:
  US13/679,890
• Time in Patent Office: 998 Days
• Field of Search: 706/12, 706/45, 706/62
• US Class Current: 1/1
• CPC Class Codes: G06F 17/11 for solving equations , e.g. nonlinear equations, general mathematical optimization problems optimization specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06F 3/0652 Erasing, e.g. deleting, data cleaning, moving of data to a wastebasket
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06N 5/047 Pattern matching networks; Rete networks

Abstract

A method is provided for estimating past data by identifying a high frequency data set for a defined time period. A pattern is calculated for the high frequency data set and then the pattern is applied to a low frequency data set in a past time period to estimate a high frequency query point.

Claims

1. A method for estimating historical metric data, comprising:

under control of a processor and memory configured with executable instructions,

identifying a plurality of fine granular metric data sets for a plurality of defined time periods which are related;

calculating a slope for each of the plurality of fine granularity metric data sets;

correlating slopes of the plurality of fine granularity metric data sets to form a correlated slope; and

applying the correlated slope to a related past time period with past coarse granularity data related to a fine granularity query for the past time period.

3. The method as in claim 1, further comprising estimating seasonality for the plurality of fine granularity data sets and fitting the correlated slope to the seasonality.

2. The method as in claim ter comprising estimating a fine granularity query point for the fine granularity query using the correlated slope inferred to the past coarse granularity data set.

4. The method as in claim 2, further comprising calculating a maximum bound limit and a minimum bound limit of the past coarse granularity data set.

5. The method as in claim 4, wherein estimating a fine granularity query point further comprises using the maximum bound limit and the minimum bound limit to estimate the fine granularity query point for the fine granularity query.

6. The method as in claim 4, wherein estimating a fine granularity query point further comprises using the maximum bound limit as a constraint on the correlated slope.

7. The method as in claim 4, wherein estimating a fine granularity query point further comprises using the minimum bound limit as a constraint on the correlated slope.

8. The method as in claim 4, wherein estimating a fine granularity query point further comprises using a mean of the maximum bound limit and the minimum bound limit as an anchor on the correlated slope.