RoBoost-PLSR :
Robust PLS regression method inspired from
boosting principles
p. 2
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Summary
1. Introduction
a. Robustness
b. Robust PLS methods
2. Theory
a. RoBoost-PLS regression
b. Pros and Cons
3. Materials and Methods
a. Methods
b. Data
4. Results and discussions
5. Conclusion
p. 3
RoBoost-PLS : robust PLS regression method inspired from boosting principles
1. Introduction
a. Robustness
p. 4
RoBoost-PLS : robust PLS regression method inspired from boosting principles
X0
1. Introduction
a. Robustness
p. 5
RoBoost-PLS : robust PLS regression method inspired from boosting principles
1. Introduction
X1
X0
p. 6
RoBoost-PLS : robust PLS regression method inspired from boosting principles
X1
X0
1. Introduction
a. Robustness
p. 7
RoBoost-PLS : robust PLS regression method inspired from boosting principles
X1
X0
1. Introduction
a. Robustness
p. 8
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Main hypothesis in Robust methods :
- Robust methods assume that the largest mass of data is X0 .
- The learning database is polluted
Main difficulties in Robust methods :
- Find a good measurement to highlight the outliers (especially when estimating
leverage points)
a. Robustness
1. Introduction
p. 9
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Wakelinc et Macfie, « A Robust PLS Procedure »
Cummins , « Iteratively reweighted partial least squares: A performance analysis
by monte carlo simulation »
Pell, « Multiple Outlier Detection for Multivariate Calibration Using Robust
Statistical Techniques »
Griep et al., « Comparison of Semirobust and Robust Partial Least Squares
Procedures »
Serneels et al., « Partial Robust M-Regression »;
Gil et Romera, « On Robust Partial Least Squares (PLS) Methods »
Møller, Frese, et Bro, « Robust Methods for Multivariate Data Analysis »
Hubert et Branden, « Robust Methods for Partial Least Squares Regression ».
b. Robust PLS methods in literature
1. Introduction
p. 10
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 11
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 12
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 13
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
PLS regression model
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 14
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
PLS regression model
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 15
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
PLS regression model
Leverage and Y-residuals
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 16
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
PLS regression model
Leverage and Y-residuals
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 17
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
PLS regression model
Leverage and Y-residuals
Weight estimation
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 18
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
PLS regression model
Leverage and Y-residuals
Weight estimation
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 19
RoBoost-PLS : robust PLS regression method inspired from boosting principles
PRM (Partial Robust M-regression) :
PLS regression model
Leverage and Y-residuals
Weight estimation
Weighting X and Y
matrices
b. Robust PLS methods in literature
1. Introduction
Until convergence
of model
Calculation of a PLS model with a defined
x latent variables then weighting according to the
p. 20
RoBoost-PLS : robust PLS regression method inspired from boosting principles
a. RoBoost-PLS
p. 21
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
a. RoBoost-PLS
2.
Theory
p. 22
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
a. RoBoost-PLS
2.
Theory
p. 23
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
a. RoBoost-PLS
2.
Theory
p. 24
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
a. RoBoost-PLS
2.
Theory
p. 25
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
Weight estimation
a. RoBoost-PLS
2.
Theory
p. 26
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
Weight estimation
a. RoBoost-PLS
2.
Theory
p. 27
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
Weight estimation
Until convergence
of model
a. RoBoost-PLS
2.
Theory
p. 28
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
Weight estimation
Until convergence
of model
a. RoBoost-PLS
2.
Theory
p. 29
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
Weight estimation
Until convergence
of model
a. RoBoost-PLS
2.
Theory
p. 30
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
Weight estimation
X = X-residuals and
Y = Y-residuals
Until convergence
of model
a. RoBoost-PLS
2.
Theory
p. 31
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
Weight estimation
X = X-residuals and
Y = Y-residuals
Until convergence
of model
a. RoBoost-PLS
2.
Theory
p. 32
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Weighted PLSR model
with 1 LV
Leverage, Y-residuals,
X-residuals
Weight estimation
X = X-residuals and
Y = Y-residuals
Until convergence
of model
up to x latent variables
a. RoBoost-PLS
2.
Theory
p. 33
RoBoost-PLS : robust PLS regression method inspired from boosting principles
- Facilitates the weighting of
leverage point
- Apprehends X-residuals
Pros
- B-coef not observable
- Scores : non-orthogonal
Cons
b. Pros and Cons
p. 34
RoBoost-PLS : robust PLS regression method inspired from boosting principles
- Facilitates the weighting of
leverage point
- Apprehends X-residuals
Pros
- B-coef not observable
- Scores : non-orthogonal
Cons
b. Pros and Cons
2.
Theory
The distance to the center of the scores
is easy to define for one dimension
p. 35
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4 different methods :
PLSR with outliers in the training set
PLSR without outliers in the training set
PRM with outliers in the training set
RoBoost-PLSR with outliers in the training set
a. Methods
p. 36
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4 different methods :
PLSR with outliers in the training set
PLSR without outliers in the training set
PRM with outliers in the training set
RoBoost-PLSR with outliers in the training set
Reference
a. Methods
p. 37
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Simulated data set
3. Materials and methods
p. 38
RoBoost-PLS : robust PLS regression method inspired from boosting principles
Real data set
3. Materials and methods
p. 39
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 40
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 41
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 42
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 43
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 44
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 45
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 46
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 47
RoBoost-PLS : robust PLS regression method inspired from boosting principles
4. Results and discussions
p. 48
RoBoost-PLS : robust PLS regression method inspired from boosting principles