study_pattern_final

what is pattern recongnition?

it's a study of how the machine observe the environment, and learn pattern
from the background and make sound and reasonable decisions about the category

what is pattern class?

is a set of patterns sharing common attributes

what is classifier?

it's a machine that performs a classification

inter class Similarty vs intra class Variability?

• inter class similarity such as twins classes, father and sons classes
  the poepel have a similarity apperence
• intra class viraibility casued by changes in light and position and three
  dimenisonal head orientation and face expressions

pattern recognition system?

1. data sensing: measure the physical variables
2. pre-pressing: remove of nosie from data
3. feature extracting: finding a new representation of data
4. model learing and estimation: learing and maping and the data into groups
5. classification: using a the features and learned the modoel to assign
   a pattern to class
6. post precessing: evaluation of confidence in decisions

pattern recognition model?

• templete model: based in similarity of objects
• satistical: based in probaliaty
• syntactic: I don't give a shit
• neural network

what is the design cycles?

• data collection
• fearture chocise
• model choice
• training
• evaluation
• computationl complexity

note

• classification: supervisied (labeled)
• clustering: Unsupervisied (unlabeled)
• study PR applications

what is bayes theorem?

Bayes theorem is extensively applied in machine learning and artificial intelligence
projects

what is fuzzy logic?

Fuzzy logic models human expertise and knowledge in some task or
application by considering all the propartites between yes or no

what is fuzzy logic2?

Fuzzy Logic (FL) is a method of reasoning that resembles human reasoning. The approach of
FL imitates the way of decision making in humans that involves all intermediate possibilities
between digital values YES and NO

what is fuzzification module?

it't transform the system input which are crisp value into fuzzy sets

what is knowledge base?

it's store if then rules provided by exportes

what is inference Engine?

it simulates the human reasoning process by making a fuzzy inference
on the input and if then rules

what is defuzzification module?

it's tranform the fuzzy set into cirsp value

what is Aggregate conclusions?

the process of unification of the outputs of all rules

what is membership functions?

Membership functions allow you to quantify linguistic term and represent
a fuzzy set graphically.

development fuzzy logic step?

• Define linguistic Variables and terms (start)
• Construct membership functions for them. (start)
• Construct knowledge base of rules (start)
• Convert crisp data into fuzzy data sets using
membership functions. (fuzzification)
• Evaluate rules in the rule base. (Inference Engine)
• Combine results from each rule. (Inference Engine)
• Convert output data into non-fuzzy values.
(defuzzification)

Differentiate between Conventional set vs. fuzzy set?

for fuzzy: the range of m(x) is [0,1]. So, the above definition cannot be used
We cannot say clearly if x is in A or not
for conventional set: the value is yes or not there clearly answer

Why Laplace estimator is used?

The Laplace estimator is a smoothing technique that adds a small constant to
observed frequencies to avoid zero probabilities,

Explain the benefit of using the Gaussian theory?

The Gaussian theory describes the behavior and properties of data that
follow a normal distribution, providing a foundation for modeling
real-world phenomena, simplifying statistical analysis
or can be describe as a model where feature vectors for given class are
continuous valued, random currpted version of a prototype vector

explain what is the benifits of using a Gaussian theory?

it's provide a mathimatically tarctable model for feature distrobutions
using bell ( curved shape)

what is laplace estimator is used in?

it's used for slove problem the zero frequency problem in classification

the components of the artificial neural networks, and describe how it works?

• inputs singal, node, synaptic weights, activation function, threshold, output
• first the inputs the singale into node and each node have weight and bais
  and making the opertion for checking if the wanted ouput or not then passing
  the output from node to activation function to handle non-linearty data
  threed check for output if is less then threethould return 0 else return 1
  then check the output from the last node to the dirserd result and return
  the error value if found

Differentiate between feedforward neural networks and recurrent neural networks?

feedforward: the data moving from the node to next one there is no checking back
for weight or any thing and might be complex(non-lineary) there is hidden layers
recurrent neural networks: more difficult to predict the ouput and the
weight update every time making a new epoch

Differentiate between feedforward backpropagation in neural networks?

feedforward: the data moving from the node to next one there is no checking back
and using the data from the node to calc intermediate for hidden layer which
is used to calc the output
backpropagation: more difficult to predict the ouput and the
weight update every time making a new !FIX

What is the main purpose of the activation function?

the main reason ot using activation function to adding non-linearty to
the neural network

Explain the Hebbian learning.

hebbinan learing happended when to nodes connected togther fire togther ,
the strength between theme increase

Explain the perceptron neural network.

Perceptron is considered a single-layer neural link with four main
parameters

Why perceptron Cannot learn exclusive-or?

because excluseive-or can represent by one lineary line