User Tools

Site Tools


how_to:computefalsepositiverate

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
how_to:computefalsepositiverate [2012/02/27 18:51]
132.168.73.124
how_to:computefalsepositiverate [2012/02/27 19:27] (current)
132.168.73.124
Line 1: Line 1:
 ====== How-to compute False Positive Rate ====== ====== How-to compute False Positive Rate ======
 +
 +See [[userguide:​falsepositiverate|here]] to have more details about False Positive Rate calculation.
  
 ===== Run the False Positive Rate algorithm ===== ===== Run the False Positive Rate algorithm =====
Line 16: Line 18:
   - Result can also be retrieved in the Context properties: \\ {{:​how_to:​falsepositiveratepropertiesgood.png|}}   - Result can also be retrieved in the Context properties: \\ {{:​how_to:​falsepositiveratepropertiesgood.png|}}
  
-===== More details about the False Positive Rate results ===== 
- 
-The Formula used to compute False Positive Rate is: 
-  * **FPR = ∑FP/​(∑TP+∑FP) ** 
-  * with FP = (2*reverse). This FP value is get from the "​significant.and.duplicated.fp"​ identification property. 
-  * with TP = (reverse+forward) - FP. This TP value is get from the "​significant.and.duplicated.tp"​ identification property. 
-  * ∑FP represents the sum of all "​significant.and.duplicated.fp"​ values (i.e. the sum of the "​significant.and.duplicated.fp"​ property of all identifications under the context) ​ 
-  * ∑TP represents the sum of all "​significant.and.duplicated.tp"​ values (i.e. the sum of the "​significant.and.duplicated.tp"​ property of all identifications under the context) ​ 
- 
-:!: **The user can retrieve the significant.and.duplicated.fp and significant.and.duplicated.tp of all the identifications in the log.** 
- 
-{{:​how_to:​falsepositiverateresultwithzero.png |}}  
- 
-If one (or more) identification contains a zero value for one or both properties, \\ 
-this identification is not taken into account in the calculation.\\ 
- 
-For example, in the screen shot on the left, you can read that 895 out of 959 identifications were used for the calculation.\\ ​ 
-\\ 
-The user is informed about this in the resulting dialog message **and** in the context properties: 
- 
-{{:​how_to:​falsepositiveratepropertieswithzero.png|}} 
- 
- 
- 
- 
- 
-===== Read the False Positive Rate result ===== 
- 
-You can see computed False Positive Rate in the Context properties: 
  
- - Select the User context for which FPR has already been computed 
- - Right-click on the selected context and select ''​Properties'' ​ 
- - The False Positive Rate appears under the "false positive"​ category: 
-   * significant.and.duplicated.rate is the rate in percentage. ​ 
  
how_to/computefalsepositiverate.1330365089.txt.gz · Last modified: 2012/02/27 18:51 by 132.168.73.124