Generating a succinct approximate representation of a time series

A method, comprising:

• at a ledger storage unit:
• recording historical ledger data, wherein the historical ledger data comprises a first sequence of actual costs incurred by a first service delivery project over time and a second sequence of actual costs incurred by a second service delivery project over time;
• at a pre-processing device including at least one hardware processor:
• retrieving the first sequence of actual costs and the second sequence of actual costs from the ledger storage unit;
• performing an operation on each sequence of actual costs retrieved to minimize noise;
• constructing a first segmented model corresponding to the first sequence of actual costs and a second segmented model corresponding to the second sequence of actual costs based on the first sequence of actual costs and the second sequence of actual costs, respectively, wherein each segmented model is maintained on a models storage unit, each segmented model has a corresponding model parameter, each segmented model comprises a plurality of segments that rise and fall alternately, each segment of each plurality of segments comprises a contiguous subsequence of a corresponding sequence of actual costs, and each plurality of segments comprises at least one segment that rises to at least one peak actual cost of a corresponding sequence of actual costs and at least one other segment that falls to at least one trough actual cost of the corresponding sequence of actual costs;
• generating a first sequence of total costs corresponding to the first segmented model and a second sequence of total costs corresponding to the second segmented model based on the first segmented model and the second segmented model, respectively, wherein each total cost of each sequence of total costs corresponds to a segment of a corresponding segmented model, each total cost comprises a sum of actual costs included in a corresponding segment, and each total cost has a corresponding length representing a number of actual costs included in a corresponding segment;
• computing a first length corresponding to the first sequence of total costs and a second length corresponding to the second sequence of total costs based on the first sequence of total costs and the second sequence of total costs, respectively, wherein each length corresponding to each sequence of total costs represents a number of total costs included in the corresponding sequence of total costs;
• and
• computing a correlation coefficient between the first sequence of total costs and the second sequence of total costs;
• at a training device including at least one hardware processor:
• for each segmented model maintained on the models storage unit, training the segmented model to better fit the historical ledger data by adjusting a corresponding model parameter to minimize a discrepancy between an aggregate of all segmented models maintained on the models storage unit and an aggregate of the historical ledger data;
• and
• training the training device to forecast long-range cost estimations corresponding to a service delivery project based on at least on segmented model maintained on the models storage unit, wherein each long-range cost estimations is maintained on a predictions storage unit, and each long-range cost estimations represents cost estimations for a corresponding service delivery project up to an end date scheduled in the future;
• and
• at a reporting device including at least one hardware processor:
• displaying an interactive user interface on a display device;
• providing, via the interactive user interface, a report indicative of whether the first service delivery project is similar to the second service delivery project, wherein the report is based on the first length and the second length computed, the correlation coefficient computed, and read-only, pre-indexed data maintained on a reporting storage unit and extracted from the first segmented model and the second segmented model maintained on the models storage unit;
• and
• reporting forecasting results corresponding to a service delivery project via the interactive user interface, wherein the forecasting results is based on the read-only, pre-indexed data maintained on the reporting storage unit and corresponding long-range cost estimations maintained on the predictions storage unit;
• wherein the historical ledger data is updated on a periodic basis.

A system, comprising:

• at least one processor;
• and
• a non-transitory processor-readable memory device storing instructions that when executed by the at least one processor causes the at least one processor to perform operations including:
• at a ledger storage unit:
• recording historical ledger data, wherein the historical ledger data comprises a first sequence of actual costs incurred by a first service delivery project over time and a second sequence of actual costs incurred by a second service delivery project over time;
• at a pre-processing device including at least one hardware processor:
• retrieving the first sequence of actual costs and the second sequence of actual costs from the ledger storage unit;
• performing an operation on each sequence of actual costs retrieved to minimize noise;
• constructing a first segmented model corresponding to the first sequence of actual costs and a second segmented model corresponding to the second sequence of actual costs based on the first sequence of actual costs and the second sequence of actual costs, respectively, wherein each segmented model is maintained on a models storage unit, each segmented model has a corresponding model parameter, each segmented model comprises a plurality of segments that rise and fall alternately, each segment of each plurality of segments comprises a contiguous subsequence of a corresponding sequence of actual costs, and each plurality of segments comprises at least one segment that rises to at least one peak actual cost of a corresponding sequence of actual costs and at least one other segment that falls to at least one trough actual cost of the corresponding sequence of actual costs;
• generating a first sequence of total costs corresponding to the first segmented model and a second sequence of total costs corresponding to the second segmented model based on the first segmented model and the second segmented model, respectively, wherein each total cost of each sequence of total costs corresponds to a segment of a corresponding segmented model, each total cost comprises a sum of actual costs included in a corresponding segment, and each total cost has a corresponding length representing a number of actual costs included in a corresponding segment;
• computing a first length corresponding to the first sequence of total costs and a second length corresponding to the second sequence of total costs based on the first sequence of total costs and the second sequence of total costs, respectively, wherein each length corresponding to each sequence of total costs represents a number of total costs included in the corresponding sequence of total costs;
• and
• computing a correlation coefficient between the first sequence of total costs and the second sequence of total costs;
• at a training device including at least one hardware processor:
• for each segmented model maintained on the models storage unit, training the segmented model to better fit the historical ledger data by adjusting a corresponding model parameter to minimize a discrepancy between an aggregate of all segmented models maintained on the models storage unit and an aggregate of the historical ledger data;
• and
• training the training device to forecast long-range cost estimations corresponding to a service delivery project based on at least on segmented model maintained on the models storage unit, wherein each long-range cost estimations is maintained on a predictions storage unit, and each long-range cost estimations represents cost estimations for a corresponding service delivery project up to an end date scheduled in the future;
• and
• at a reporting device including at least one hardware processor:
• displaying an interactive user interface on a display device;
• providing, via the interactive user interface, a report indicative of whether the first service delivery project is similar to the second service delivery project, wherein the report is based on the first length and the second length computed, the correlation coefficient computed, and read-only, pre-indexed data maintained on a reporting storage unit and extracted from the first segmented model and the second segmented model maintained on the models storage unit;
• and
• reporting forecasting results corresponding to a service delivery project via the interactive user interface, wherein the forecasting results is based on the read-only, pre-indexed data maintained on the reporting storage unit and corresponding long-range cost estimations maintained on the predictions storage unit;
• wherein the historical ledger data is updated on a periodic basis.

A non-transitory computer program product for analyzing and comparing service delivery projects, the computer program product comprising a tangible storage medium readable by a computer system and storing instructions for execution by the computer system for performing a method comprising:

• at a ledger storage unit:
• recording historical ledger data, wherein the historical ledger data comprises a first sequence of actual costs incurred by a first service delivery project over time and a second sequence of actual costs incurred by a second service delivery project over time;
• at a pre-processing device including at least one hardware processor:
• retrieving the first sequence of actual costs and the second sequence of actual costs from the ledger storage unit;
• performing an operation on each sequence of actual costs retrieved to minimize noise;
• constructing a first segmented model corresponding to the first sequence of actual costs and a second segmented model corresponding to the second sequence of actual costs based on the first sequence of actual costs and the second sequence of actual costs, respectively, wherein each segmented model is maintained on a models storage unit, each segmented model has a corresponding model parameter, each segmented model comprises a plurality of segments that rise and fall alternately, each segment of each plurality of segments comprises a contiguous subsequence of a corresponding sequence of actual costs, and each plurality of segments comprises at least one segment that rises to at least one peak actual cost of a corresponding sequence of actual costs and at least one other segment that falls to at least one trough actual cost of the corresponding sequence of actual costs;
• generating a first sequence of total costs corresponding to the first segmented model and a second sequence of total costs corresponding to the second segmented model based on the first segmented model and the second segmented model, respectively, wherein each total cost of each sequence of total costs corresponds to a segment of a corresponding segmented model, each total cost comprises a sum of actual costs included in a corresponding segment, and each total cost has a corresponding length representing a number of actual costs included in a corresponding segment;
• computing a first length corresponding to the first sequence of total costs and a second length corresponding to the second sequence of total costs based on the first sequence of total costs and the second sequence of total costs, respectively, wherein each length corresponding to each sequence of total costs represents a number of total costs included in the corresponding sequence of total costs;
• and
• computing a correlation coefficient between the first sequence of total costs and the second sequence of total costs;
• at a training device including at least one hardware processor:
• for each segmented model maintained on the models storage unit, training the segmented model to better fit the historical ledger data by adjusting a corresponding model parameter to minimize a discrepancy between an aggregate of all segmented models maintained on the models storage unit and an aggregate of the historical ledger data;
• and
• training the training device to forecast long-range cost estimations corresponding to a service delivery project based on at least on segmented model maintained on the models storage unit, wherein each long-range cost estimations is maintained on a predictions storage unit, and each long-range cost estimations represents cost estimations for a corresponding service delivery project up to an end date scheduled in the future;
• and
• at a reporting device including at least one hardware processor:
• displaying an interactive user interface on a display device;
• providing, via the interactive user interface, a report indicative of whether the first service delivery project is similar to the second service delivery project, wherein the report is based on the first length and the second length computed, the correlation coefficient computed, and read-only, pre-indexed data maintained on a reporting storage unit and extracted from the first segmented model and the second segmented model maintained on the models storage unit;
• and
• reporting forecasting results corresponding to a service delivery project via the interactive user interface, wherein the forecasting results is based on the read-only, pre-indexed data maintained on the reporting storage unit and corresponding long-range cost estimations maintained on the predictions storage unit;
• wherein the historical ledger data is updated on a periodic basis.