2. PLS STATISTICAL
TECHNIQUE
PLS STATISTICAL
TECHNIQUE
PLS STATISTICAL
TECHNIQUE
Partial Least Squares (PLS) is a type of regression analysis that
is used to analyze relationships between a large number of
predictor variables and one or more response variables.
01 Used for predictive modelling and
dimension reduction of large datasets
02
Similar to principal component analysis (PCA)
but unlike PCA, it can handle both independent
and dependent variables
03
Supervised learning algorithm, meaning it
relies on labeled data (i.e. a response variable)
to build its model
04
Reduce the complexity of large datasets,
identify important relationships between
variables, and predict outcomes
3. PLS IN
CONSUMER
ANALYTICS
It can be used in consumer analytics to explore
relationships between a large set of variables, such as
consumer behaviors, preferences, demographics, and
product features. PLS is particularly useful when there
are many variables, and some of them may be highly
correlated, making it difficult to isolate the effects of
individual factors.
PLS IN
CONSUMER
ANALYTICS
4. PROS
PLS is a nonparametric technique and does
not require any assumptions about the
distribution of the data or the relationships
between variables. Moreover it is easy to
interpret and helps to identify important
relationships between variables in the data.
01
02
03
04
05
Powerful predictive modeling technique that
can handle high levels of multicollinearity
among predictor variables
Robust method and can often produce more
accurate predictions than traditional
regression methods
Can be used to handle nonlinear relationships
between the predictors and outcomes
Can handle larger datasets with more
variables than traditional regression methods
Requires fewer assumptions and does not
require the data to be normally distributed
PROS
5. CONS
PLS is highly sensitive to outliers. Outliers, or
points that are far away from the rest of the
data, can significantly skew the results of a
PLS analysis. It is important to consider the
type of data being analyzed and the desired
results before deciding if PLS is the best
method for the job.
01
02
03
04
More difficult to interpret than traditional
regression methods because it relies on latent
variables that are not directly observed
More computationally intensive than
traditional regression methods and may
require more time to run
May be less accurate than traditional regression
methods when there is a low degree of
multicollinearity
May produce inaccurate results when the
number of predictor variables is greater than
the sample size
CONS
6. APPLICATIONS
OF PARTIAL
LEAST SQUARE
MODELS
Partial least square models can be used for a
variety of applications in consumer analytics.
They can be used to identify customer segments,
predict customer behavior, and analyze
customer journeys.
The models can also be used to optimize
marketing campaigns, improve customer
retention, and identify new opportunities for
growth.
APPLICATIONS
OF PARTIAL
LEAST SQUARE
MODELS
8. Limitations of Partial Least Square
Models
Sample Size
Model Complexity
Data Quality
Interpretability
Limited to linear relationships
Generalizability
Partial least square models are not without their limitations. The models can be computationally intensive, and
require significant amounts of data to be accurate.
The models can also be difficult to interpret, as the relationships between variables can be complex and difficult
to understand.
Some limitations of PLS Models in Consumer Analytics are as follows:
1.
2.
3.
4.
5.
6.
9. References
Henseler, J., Ringle, C. M., & Sinkovics, R. R. (2009). The use of partial least squares path modeling in
international marketing. In New challenges to international marketing. Emerald Group Publishing Limited.
http://www.scholarpedia.org/article/Partial_differential_equation#:~:text=Partial%20differential%20equati
ons%20are%20used,%2C%20electrostatics%2C%20electrodynamics%2C%20etc.
https://www.xlstat.com/en/solutions/features/partial-least-squares-
regression#:~:text=The%20Partial%20Least%20Squares%20regression,used%20to%20perfom%20a%20re
gression.
https://www.ibm.com/docs/en/spss-statistics/27.0.0?topic=features-partial-least-squares-regression
https://www.investopedia.com/terms/l/least-squares-method.asp
https://community.tibco.com/s/article/TIBCO-Statistica-General-Partial-Least-Squares-Models