Automated trend detection by self-learning models through image generation and recognition

A computer program product comprising:

• a computer readable storage medium readable by one or more processors and storing instructions for execution by the one or more processors for performing a method comprising:
• monitoring, by the one or more processors, in real-time, resources comprising a computing system to identify events in the computing system and timing of the events, wherein the events and the timing of the events over a pre-defined period of time comprise time series trend data, wherein the pre-defined period of time comprises more than one vintage, wherein each vintage of the more than one vintage comprises a collection of contemporaneous events, wherein a portion of the events comprise system failures;
• obtaining, by the one or more processors, the time series trend data;
• transforming, by the one or more processors, the time series trend data into measurable dimensions, wherein the transforming further comprises assigning the transformed data to vintages of the more than one vintage and to distinct periods of time, wherein the transforming comprises generating, by the one or more processors, pixels as data indices representing the more than one vintage;
• and
• generating, by the one or more processors, from the transformed data assigned to the vintages, an image, wherein the image represents trends in the time series trend data, organized by the distinct periods of times and the vintages, wherein each pixel represents a presence of an event or an absence of the event of the monitored events comprising the time series trend data, and wherein a color of each pixel representing the presence of an event indicates a severity of the event, wherein the image comprises at least one row of pixels and at least one column of pixels, wherein each column of the at least one column represents a distinct period of time of the distinct periods of time, and wherein each row of the at least one row represents a vintage of the more than one vintage.

A system comprising:

• a memory;
• one or more processors in communication with the memory;
• program instructions executable by the one or more processors via the memory to perform a method, the method comprising:
• monitoring, by the one or more processors, in real-time, resources comprising a computing system to identify events in the computing system and timing of the events, wherein the events and the timing of the events over a pre-defined period of time comprise time series trend data, wherein the pre-defined period of time comprises more than one vintage, wherein each vintage of the more than one vintage comprises a collection of contemporaneous events, wherein a portion of the events comprise system failures;
• obtaining, by the one or more processors, the time series trend data;
• transforming, by the one or more processors, the time series trend data into measurable dimensions, wherein the transforming further comprises assigning the transformed data to vintages of the more than one vintage and to distinct periods of time, wherein the transforming comprises generating, by the one or more processors, pixels as data indices representing the more than one vintage;
• and
• generating, by the one or more processors, from the transformed data assigned to the vintages, an image, wherein the image represents trends in the time series trend data, organized by the distinct periods of times and the vintages, wherein each pixel represents a presence of an event or an absence of the event of the monitored events comprising the time series trend data, and wherein a color of each pixel representing the presence of an event indicates a severity of the event, wherein the image comprises at least one row of pixels and at least one column of pixels, wherein each column of the at least one column represents a distinct period of time of the distinct periods of time, and wherein each row of the at least one row represents a vintage of the more than one vintage.