ToolboxPerformance: Difference between revisions

From Eigenvector Research Documentation Wiki
Jump to navigation Jump to search
imported>Mathias
imported>Mathias
Line 16: Line 16:


'''Table 1. Properties of different cross-validation methods in Solo and PLS_Toolbox.'''
'''Table 1. Properties of different cross-validation methods in Solo and PLS_Toolbox.'''
{| border="1" cellpadding="5" cellspacing="0"
| ||'''Venetian Blinds'''||'''Contiguous Blocks'''||'''Random Subsets'''||'''Leave-One Out'''||'''Custom'''
|- valign="top"
| |
||
|| 
|| 
|| 
||
|- valign="top"
| |
||
||
||
||
||
|- valign="top"
| |
||
||
||
||
||
|-
|
||
||
||
||
||
|-
|}
'''Table 2. Performance of nnon-linear methods'''


{| border="1" cellpadding="5" cellspacing="0"
{| border="1" cellpadding="5" cellspacing="0"
Line 24: Line 95:
| |'''Test sample selection scheme'''
| |'''Test sample selection scheme'''
||
||
[[Image:Cv_vet.jpg||| ]]
 
||   
||   
[[Image:Cv_con.jpg||| ]]
 
||   
||   
[[Image:Cv_rnd.jpg||| ]]
 
||   
||   
[[Image:Cv_loo.jpg||| ]]
 
||
||
* User-defined subsets
 
* Can "force" specific objects into every test set, every model set, or exclude them from the CV procedure
 
|- valign="top"
| |'''100 Samples''
||
 
||
 
||
 
||
 
||
 
|- valign="top"  
| |'''500 Samples'''
||
 
||
 
||
 
 
||
 
||
 
 
|-
 
|'''1000 samples'''
||
 
||
 
||
 
||
 
||
 
 
|-
 
|}
 
 
 
 
 
 
 
'''Table 2. Performance of nnon-linear methods'''
 
{| border="1" cellpadding="5" cellspacing="0"
 
| ||'''Venetian Blinds'''||'''Contiguous Blocks'''||'''Random Subsets'''
 
|- valign="top"
| |'''Test sample selection scheme'''
||
 
|| 
 
|| 
 
 
 
|- valign="top"  
|- valign="top"  
| |'''Parameters'''
| |'''Parameters'''
||
||
* Number of Data Splits (s)
 
* Maximum number of PCs/LVs
* Total number of objects/samples (n)
||
||
* Number of Data Splits (s)
 
* Maximum number of PCs/LVs
* Total number of objects/samples (n)
||
||
* Number of Data Splits (s)
 
* Number of iterations (r)
 
* Maximum number of PCs/LVs
 
* Total number of objects/samples (n)
|- valign="top"
| |'''Number of sub-validation experiments'''
||
||
* Maximum number of PCs/LVs
 
* Total number of objects/samples (n)
||
||
* Number of data splits (s)
 
* Object membership for each split
||
* All user-defined
 
* Total number of objects/samples (n)
 
 
 
|-
 
 
 
|}
 
 
 
 
'''Table 3. ANN'''
 
{| border="1" cellpadding="5" cellspacing="0"
 
| ||'''Venetian Blinds'''||'''Contiguous Blocks'''||'''Random Subsets'''
 
|- valign="top"  
|- valign="top"  
| |'''Number of sub-validation experiments'''
| |'''Test sample selection scheme'''
||
 
|| 
 
|| 
 
 
 
|- valign="top"
| |'''Parameters'''
||
 
||
||
= s
 
||
||
= s
 
 
 
|- valign="top"
| |'''Number of sub-validation experiments'''
||
||
= (s * r)


||
||
= n
 
||
||
= s
 
 
 


|-
|-


|'''Number of test samples per sub-validation'''
 
 
|}
 
 
 
 
 
'''Table 4. SVM'''
 
{| border="1" cellpadding="5" cellspacing="0"
 
| ||'''100 variables'''||'''500 variables'''||'''1000 variables'''
 
|- valign="top"
| |'''100 samples'''
||
 
|| 
 
|| 
 
 
 
|- valign="top"
| |'''500 samples'''
||
 
||
||
= n/s
 
||
||
= n/s
 
 
 
|- valign="top"
| |'''1000 samples'''
||
||
= n/s
 
||
||
=1
 
||
||
* Can vary, user defined
 
 
 


|-
|-


|}
|}

Revision as of 13:22, 8 September 2016

PLS_Toolbox Performance

The following performance results are for general comparison and expectation. Your own mileage may vary.


Performance Table
Matlab Versoin PLS_Toolbox Version Operating System System Description Data Description Algorithm Performance Result
2015a 8.1.1 OS X El Capitan 2.8 GHz Intel, 16 GB ram cell cell



Table 1. Properties of different cross-validation methods in Solo and PLS_Toolbox.

Venetian Blinds Contiguous Blocks Random Subsets Leave-One Out Custom









Table 2. Performance of nnon-linear methods

Venetian Blinds Contiguous Blocks Random Subsets Leave-One Out Custom
Test sample selection scheme


'100 Samples
500 Samples



1000 samples





Table 2. Performance of nnon-linear methods

Venetian Blinds Contiguous Blocks Random Subsets
Test sample selection scheme


Parameters


Number of sub-validation experiments





Table 3. ANN

Venetian Blinds Contiguous Blocks Random Subsets
Test sample selection scheme


Parameters


Number of sub-validation experiments





Table 4. SVM

100 variables 500 variables 1000 variables
100 samples


500 samples


1000 samples