|
|  |  | Observation classification, estimate of the subject state and suggestion of the diagnosis | | | Aim | » | To increase results accuracy, to exclude errors connected with human element | | | Premises | » | Suppose you've conducted an experiment and obtained several numerical characteristics of the object (e.g., test resuts for some product qualities or for condition of the patient). Suppose you know that that object can belong to one of the classes (healthy/ill, functions properly, has malfunction ¹1, malfunction ¹2 etc.). And finally suppose you possess cumulative information about former tests for similar objects. Then it is necessary to give most reliable estimation or diagnosis for state of the object using the previous tests results. | | | Our suggestion | » | We suggest construction of the model which will permit to make decisions about the objects belonging to one of the classes automatically. We'll also prepare a report comprising - main rules permitting to make decision,
- accuracy of the constructed rules,
- method of the model construction,
- found errors in measuring results,
- conclusions and recommendations.
| | | | | We'll consider with interest any of your suggestions | | If you've got any questions then leave your message. | |
| |
|
|
