Method for automatically identifying signals or patterns in time series data by treating series as image

Abstract

A method for analyzing time-series data is provided. The method includes: identifying, from within a first set of time-series data, data sequences that correspond to patterns, each data sequence being representative of a predetermined time interval; converting the data sequences into respective pattern images; receiving a second set of time-series data; converting the second set of time-series data into an additional image; comparing each of the pattern images with a portion of the additional image that corresponds to a most recent time interval; and determining whether the portion of the additional image that corresponds to the most recent time interval corresponds to any of the patterns based on a result of the comparison.

What is claimed is: 
     
         1 . A method for analyzing time-series data, comprising:
 identifying, from within a first set of time-series data, at least a first data sequence that corresponds to a first pattern, and at least a second data sequence that corresponds to a second pattern, each of the first data sequence and the second data sequence being representative of a predetermined time interval;   converting the first data sequence into a first image and converting the second data sequence into a second image;   receiving a second set of time-series data;   converting the second set of time-series data into a third image;   comparing each of the first image and the second image with a portion of the third image that corresponds to a most recent time interval; and   determining whether the portion of the third image that corresponds to the most recent time interval corresponds to at least one of the first pattern and the second pattern based on a result of the comparison.   
     
     
         2 . The method of  claim 1 , further comprising augmenting each of the first image and the second image with at least one box-and-whisker notation that indicates an opening value, a highest value, a lowest value, a closing value, and an indication regarding whether the opening value is greater than or less than the closing value for each respective image. 
     
     
         3 . The method of  claim 2 , wherein the determining comprises using the augmented first image and the augmented second image to determine whether the corresponding at least one of the first pattern and the second pattern corresponds to at least one of an indication for a buying opportunity and a selling opportunity, and when the buying opportunity is indicated, the method further comprises transmitting a message that includes a buy recommendation to a predetermined destination, and when the selling opportunity is indicated, the method further comprises transmitting a message that includes a sell recommendation to the predetermined destination. 
     
     
         4 . The method of  claim 3 , wherein the determining further comprises using an algorithm that implements a Bollinger Bands (BB) crossing technique for determining whether at least one of a buying opportunity and a selling opportunity is indicated. 
     
     
         5 . The method of  claim 3 , wherein the determining further comprises using an algorithm that implements a Moving Average Convergence Divergence (MACD) crossing technique for determining whether at least one of a buying opportunity and a selling opportunity is indicated. 
     
     
         6 . The method of  claim 3 , wherein the determining further comprises using an algorithm that implements a Relative Strength Index (RSI) crossing technique for determining whether at least one of a buying opportunity and a selling opportunity is indicated. 
     
     
         7 . The method of  claim 3 , wherein the converting the second set of time-series data into the third image comprises implementing a 30-day window with respect to the third image. 
     
     
         8 . The method of  claim 2 , wherein the augmenting further comprises linearly varying a respective width of each of the at least one box-and-whisker notation. 
     
     
         9 . The method of  claim 2 , wherein the augmenting further comprises varying a respective width of each of the at least one box-and-whisker notation based on a corresponding change in trade volume. 
     
     
         10 . The method of  claim 2 , wherein the augmenting further comprises overlaying a previous respective closing value on each of the at least one box-and-whisker notation. 
     
     
         11 . The method of  claim 2 , wherein the determining comprises using the augmented first image and the augmented second image to determine whether the corresponding at least one of the first pattern and the second pattern corresponds to at least one of an indication for a volatile market and a calm market, and when the volatile market is indicated, the method further comprises transmitting a message that includes information that relates to market volatility to a predetermined destination, and when the calm market is indicated, the method further comprises transmitting a message that includes information that relates to market calmness to the predetermined destination. 
     
     
         12 . The method of  claim 2 , wherein the determining comprises using the augmented first image and the augmented second image to determine whether the corresponding at least one of the first pattern and the second pattern corresponds to at least one of an indication for an economic growth and a recession, and when the economic growth is indicated, the method further comprises transmitting a message that includes information that relates to economic growth to a predetermined destination, and when the recession is indicated, the method further comprises transmitting a message that includes information that relates to the recession to the predetermined destination. 
     
     
         13 . A computing apparatus for analyzing time-series data, the computing apparatus comprising:
 a processor;   a memory; and   a communication interface coupled to each of the processor, the memory, and a display screen,   wherein the processor is configured to:
 identify, from within a first set of time-series data, at least a first data sequence that corresponds to a first pattern, and at least a second data sequence that corresponds to a second pattern, each of the first data sequence and the second data sequence being representative of a predetermined time interval; 
 convert the first data sequence into a first image and convert the second data sequence into a second image; 
 receive, via the communication interface, a second set of time-series data; 
 convert the second set of time-series data into a third image; 
 compare each of the first image and the second image with a portion of the third image that corresponds to a most recent time interval; and 
 determine whether the portion of the third image that corresponds to the most recent time interval corresponds to at least one of the first pattern and the second pattern based on a result of the comparison. 
   
     
     
         14 . The computing apparatus of  claim 13 , wherein the processor is further configured to augment each of the first image and the second image with at least one box-and-whisker notation that indicates an opening value, a highest value, a lowest value, a closing value, and an indication regarding whether the opening value is greater than or less than the closing value for each respective image. 
     
     
         15 . The computing apparatus of  claim 14 , wherein the processor is further configured to use the augmented first image and the augmented second image to determine whether the corresponding at least one of the first pattern and the second pattern corresponds to an indication for at least one of an indication for a buying opportunity and a selling opportunity, and when the buying opportunity is indicated, to transmit a message that includes a buy recommendation to a predetermined destination, and when the selling opportunity is indicated, to transmit a message that includes a sell recommendation to the predetermined destination. 
     
     
         16 . The computing apparatus of  claim 15 , wherein the processor is further configured to use an algorithm that implements a Bollinger Bands (BB) crossing technique for determining whether at least one of a buying opportunity and a selling opportunity is indicated. 
     
     
         17 . The computing apparatus of  claim 15 , wherein the processor is further configured to use an algorithm that implements a Moving Average Convergence Divergence (MACD) crossing technique for determining whether at least one of a buying opportunity and a selling opportunity is indicated. 
     
     
         18 . The computing apparatus of  claim 15 , wherein the processor is further configured to use an algorithm that implements a Relative Strength Index (RSI) crossing technique for determining whether at least one of a buying opportunity and a selling opportunity is indicated. 
     
     
         19 . The computing apparatus of  claim 15 , wherein the processor is further configured to implement a 30-day window with respect to the third image. 
     
     
         20 . The computing apparatus of  claim 14 , wherein the processor is further configured to linearly vary a respective width of each of the at least one box-and-whisker notation. 
     
     
         21 . The computing apparatus of  claim 14 , wherein the processor is further configured to vary a respective width of each of the at least one box-and-whisker notation based on a corresponding change in trade volume. 
     
     
         22 . The computing apparatus of  claim 14 , wherein the processor is further configured to overlay a previous respective closing value on each of the at least one box-and-whisker notation. 
     
     
         23 . The computing apparatus of  claim 14 , wherein the processor is further configured to use the augmented first image and the augmented second image to determine whether the corresponding at least one of the first pattern and the second pattern corresponds to at least one of an indication for a volatile market and a calm market, and when the volatile market is indicated, to transmit a message that includes information that relates to market volatility to a predetermined destination, and when the calm market is indicated, to transmit a message that includes information that relates to market calmness to the predetermined destination. 
     
     
         24 . The computing apparatus of  claim 14 , wherein the processor is further configured to use the augmented first image and the augmented second image to determine whether the corresponding at least one of the first pattern and the second pattern corresponds to at least one of an indication for an economic growth and a recession, and when the economic growth is indicated, to transmit a message that includes information that relates to economic growth to a predetermined destination, and when the recession is indicated, to transmit a message that includes information that relates to the recession to the predetermined destination.