4. • Everyday actions are
increasingly being handled
electronically, instead of
pencil and paper or face to
face.
• It is useful in day by day life.
For instance, There are
several different places which
are used cameras for different
purposes.
11. • Step 1: Collect the database from
different persons for face expression
using MATLAB and collect the raw
data into excel sheet.
• Step 2: Pre-process and filter the
collected raw data into formatted format
and also apply filtration techniques like
replace missing values etc.
• Step 3: Classification using naïve
bayes, KNN, MLP, Decision tree using
Weka tool.
12. Naïve Bayes Algorithm
• It is based on Bayes' Theorem.
Naïve Bayes classifier can predict
class membership probabilities
such as the probability that a
given tuple belongs to a particular
class.
14. Decision Tree Algorithm
• A decision tree is a structure
that includes a root node,
branches, and leaf nodes.
The topmost node in the tree
is the root node.
16. Multi-Layer Perceptron (MLP)
Algorithm
• A multilayer perceptron (MLP) is
a class of artificial neural network.
An MLP consists of at least three
layers of nodes.
• Input Layer: This layer generates the input for
the network.
• Hidden Layer: The layer that maps the input to
the corresponding output is named as a hidden
layer.
• Output Layer: The layer from where the
resultant can be seen.
18. K-Nearest Neighbor (k-NN) Algorithm
• k-NN is a type of instance-based
learning, or lazy learning, is
a non-parametric method used
for classification and regression.
Both for classification and
regression, a useful technique can
be to assign weight to the
contributions of the neighbors, so
that the nearer neighbors
contribute more to the average
than the more distant ones.
32. • In this project, facial emotions
classification is done using Machine
Learning Classifiers. Each facial
expression features are extracted and
categorized into 7 categories: Happy, Sad,
Angry, Disgust, Surprise, Neutral and
Fear.
33. Future Work…
• In future, the performance of
KNN algorithm can be improved
further and also the analysis can
be done on speech and another
dataset.