Project predictor

1. A computer implemented method for project prediction, comprising:

obtaining historical project data of historical projects;

analyzing, with an analyzer stored as a set of computer readable instructions on a computer readable storage media and executable by a processor, the historical project data to generate at least one model for predicting a proposed project cycle time, wherein the analyzing to generate the at least one model comprises:

identifying at least one transition data point in the historical project data corresponding to a difference between actual cycle times for the historical projects that is indicative of a difference in the nature of the historical projects on either side of the at least one transition data point, wherein identifying the at least one transition data point comprises arranging projects in order based on actual cycle times for the historical projects, where the difference between actual cycle times for the historical projects corresponds to a transition data point where the actual cycle times transition from a first trend in the actual cycle times to a second trend in the actual cycle times;

dividing the historical project data into historical data ranges, one historical data range approximately on either side of the at least one transition data point; and

generating a linear sub-model for predicting project cycle time for each of the historical data ranges using multiple regression analysis on characteristics of the historical projects in the historical data range;

selecting the at least one model for predicting the proposed project cycle time, wherein the selected model comprises the linear sub-models corresponding to the historical data ranges; and

applying proposed project data to one of the linear sub-models to predict the proposed project cycle time, wherein a proposed data range of the proposed project corresponds with the historical data range corresponding with the one of the linear sub-models.

2. The computer implemented method of claim 1 wherein the proposed project data comprises proposed characteristics, wherein the historical project data comprises historical cycle times and historical characteristics significantly correlated with the historical cycle times, and wherein the proposed characteristics and the historical characteristics comprise at least one of estimated costs, organization participation, a number of applications effected, a project size, and a pre-existing project status.

3. The computer implemented method of claim 1 wherein selecting the at least one model for the proposed project cycle time comprises testing models generated by the analyzing of the historical project data by at least one of a root mean square error method, a coefficient of determination method, an adjusted coefficient of determination method, and an empirical method.

4. The computer implemented method of claim 1 wherein the proposed data range is based on an estimated total cost.

5. A computer implemented method for project prediction, comprising:

obtaining historical project data of historical projects;

analyzing the historical project data to generate at least one model for predicting a proposed project cycle time, wherein the analyzing to generate the at least one model comprises:

identifying at least one transition data point in the historical project data corresponding to a difference between actual cycle times for the historical projects that is indicative of a difference in the nature of the historical projects on either side of the at least one transition data point;

dividing the historical project data into historical data ranges, one historical data range approximately on either side of the at least one transition data point; and

generating a linear sub-model for predicting project cycle time for each of the historical data ranges using multiple regression analysis on characteristics of the historical projects in the historical data range;

selecting the at least one model for predicting the proposed project cycle time, wherein the selected model comprises the linear sub-models corresponding to the historical data ranges;

applying proposed project data to one of the linear sub-models to predict the proposed project cycle time, wherein a proposed data range of the proposed project corresponds with the historical data range corresponding with the one of the linear sub-models; and

updating the selected model using the proposed project data upon completing the proposed project;

wherein the at least one transition data point is a specific estimated cost; and

wherein the linear sub-model generated using multiple regression analysis for the historical data range that is greater than about the specific estimated cost based on characteristics of the historical projects including a project size and organization participation.

6. A computer implemented method for project prediction, comprising:

obtaining historical project data of historical projects;

analyzing the historical project data to generate at least one model for predicting a proposed project cycle time, wherein the analyzing to generate the at least one model comprises:

identifying at least one transition data point in the historical project data corresponding to a difference between actual cycle times for the historical projects that is indicative of a difference in the nature of the historical projects on either side of the at least one transition data point;

dividing the historical project data into historical data ranges, one historical data range approximately on either side of the at least one transition data point; and

generating a linear sub-model for predicting project cycle time for each of the historical data ranges using multiple regression analysis on characteristics of the historical projects in the historical data range;

selecting the at least one model for predicting the proposed project cycle time, wherein the selected model comprises the linear sub-models corresponding to the historical data ranges;

applying proposed project data to one of the linear sub-models to predict the proposed project cycle time, wherein a proposed data range of the proposed project corresponds with the historical data range corresponding with the one of the linear sub-models; and

updating the selected model using the proposed project data upon completing the proposed project;

wherein the at least one transition data point is a specific estimated cost; and

wherein the linear sub-model generated using multiple regression analysis for the historical data range that is less than about the specific estimated cost based on characteristics of the historical projects including a number of applications affected and a pre-existing project status.

7. A computer implemented system for project prediction, comprising:

a data manager stored on a computer readable storage medium, which when executed by a processor obtains historical project data;

an analyzer stored on a computer readable storage medium, which when executed by a processor analyzes the historical project data to generate at least one model for predicting a proposed project cycle time, wherein the analysis to generate the at least one model comprises:

identifying at least one transition data point in the historical project data corresponding to a difference between actual cycle times for the historical projects that is indicative of a difference in the nature of the historical projects on either side of the at least one transition data point wherein the transition data point is based on a cost threshold;

dividing the historical project data into historical data ranges, one historical data range approximately on either side of the at least one transition data point; and

generating a linear sub-model for predicting project cycle time for each of the historical data ranges using multiple regression analysis on characteristics of the historical projects in the historical data range; and

a user interface stored on a computer readable storage medium, which when executed by a processor:

selects the at least one model for predicting the proposed project cycle time, wherein the selected model comprises the linear sub-models corresponding to the historical data ranges, and

applies proposed project data to one of the linear sub-models to predict the proposed project cycle time, wherein a proposed data range of the proposed project corresponds with the historical data range corresponding with the one of the linear sub-models.

8. The computer implemented system of claim 7 wherein the proposed project data comprises proposed characteristics, and wherein the historical project data comprises historical cycle times and historical characteristics significantly correlated with the historical cycle times, and wherein the proposed characteristics and the historical characteristics comprise at least one of estimated costs, organization participation, a number of applications affected, a project size, and a pre-existing project status.

9. The computer implemented system of claim 7 wherein the user interface is further operable to test models generated by the analysis of the historical project data by at least one of a root mean square error method, a coefficient of determination method, an adjusted coefficient of determination method, and an empirical method.

10. A computer implemented method for project prediction, comprising:

obtaining historical project data of historical projects;

analyzing with an analyzer stored as a set of computer readable instructions on a computer readable storage media and executable by a processor the historical project data to generate at least one model for predicting a proposed project cost, wherein the analyzing to generate the at least one model comprises:

identifying at least one transition data point in the historical project data corresponding to a difference between actual project cost for the historical projects that is indicative of a difference in the nature of the historical projects on either side of the at least one transition data point, wherein identifying the at least one transition data point comprises arranging projects in order based on actual cycle times for the historical projects, where the difference between actual cycle times for the historical projects corresponds to a transition data point where the actual cycle times transition from a first trend in the actual cycle times to a second trend in the actual cycle times;

dividing the historical project data into historical data ranges, one historical data range approximately on either side of the at least one transition data point; and

generating a linear sub-model for predicting project cycle time for each of the historical data ranges using multiple regression analysis on characteristics of the historical projects in the historical data range;

selecting the at least one model for predicting the proposed project cost, wherein the selected model comprises the linear sub-models corresponding to the historical data ranges;

applying proposed project data to one of the linear sub-models to predict the proposed project cost, wherein a proposed data range of the proposed project corresponds with the historical data range corresponding with the one of the linear sub-models; and

updating the selected model using the proposed project data upon completing the proposed project.

11. The computer implemented method of claim 10 wherein the proposed project data comprises proposed characteristics, and wherein the historical project data comprises historical costs and historical characteristics significantly correlated with the historical costs.

12. The computer implemented method of claim 10 wherein the proposed characteristics and the historical characteristics comprise at least one of estimated costs, organization participation, a number of applications affected, a project size, and a pre-existing project status.

13. The computer implemented method of claim 7 wherein selecting the at least one model for the proposed project cost comprises testing models generated by the analyzing of the historical project data by at least one of a root mean square error method, a coefficient of determination method, and an adjusted coefficient of determination method, and an empirical method.

14. The computer implemented method of claim 1 wherein one of the linear sub-models generated using multiple regression analysis for one the historical data ranges on one side of the transition data point is based on a first set of characteristics of the historical projects, and the other of the linear sub-models generated using multiple regression analysis for the other of the historical data ranges on the other side of the transition data point is based on a second set of characteristics of the historical projects.

15. The computer implemented system of claim 7 wherein one of the linear sub-models generated using multiple regression analysis for one the historical data ranges on one side of the transition data point is based on a first set of characteristics of the historical projects, and the other of the linear sub-models generated using multiple regression analysis for the other of the historical data ranges on the other side of the transition data point is based on a second set of characteristics of the historical projects.

16. The computer implemented method of claim 10 wherein one of the linear sub-models generated using multiple regression analysis for one the historical data ranges on one side of the transition data point is based on a first set of characteristics of the historical projects, and the other of the linear sub-models generated using multiple regression analysis for the other of the historical data ranges on the other side of the transition data point is based on a second set of characteristics of the historical projects.

17. The computer implemented method of claim 1 , wherein the point in the arranged historical projects where the actual cycle times transition from the first trend in the actual cycle times to the second trend in the actual cycle times is based on comparing an average change between actual cycle times for successive historical projects in the arranged historical projects.

18. The computer implemented method of claim 1 , wherein identifying the at least one transition data point further comprises identifying at least one characteristic of the historical projects with values that when compared with the actual cycle times of the historical project about the transition cycle time are indicative of the difference in the nature of the historical projects.

19. The computer implemented method of claim 18 , wherein the at least one transition data point is a value of the at least one characteristic when compared with the actual cycle times of the historical projects that corresponds with the transition cycle time.

20. The computer implemented method of claim 19 , wherein the at least one characteristic is project cost.