# Cognitive Radio Questions and Answers – Techniques – Artificial Intelligence – 3

«
»

This set of Cognitive Radio Quiz focuses on “Techniques – Artificial Intelligence – 3”.

1. Which among the following gives the definition of fuzzy logic?
a) Fuzzy logic follows Boolean logic where zero stands for true and one stands for false
b) Fuzzy logic follows Boolean logic where zero stands for false and one stands for true
c) Fuzzy logic allows the assignment of any real number from zero to one to variables
d) Fuzzy logic allows the assignment of any whole number from zero to one to variables

Explanation: Fuzzy logic allows the assignment of any real number from zero to one to variables. It is a multi-valued logic. It implements the condition of partial truth. The truth value may fall anywhere between completely false to completely true.

2. Which among the following is a false statement about Fuzzy logic?
a) Fuzzy logic may use imprecise information
b) Fuzzy logic uses only accurate information
c) Fuzzy logic may represent vague information
d) Fuzzy logic may use non-numerical information

Explanation: Fuzzy logic is based on the observation that decision making sometimes involves making use of inaccurate and imprecise information. The fuzzy logic models are generally vague but are capable of identifying, representing, and modifying unclear information and data.

3. Which among the following is the correct sequence of events that occur in application of fuzzy logic?
a) Fuzzify the input, execute the applicable rules, and de-fuzzify the output
b) Execute the applicable rules, fuzzify the input, and de-fuzzify the output
c) Fuzzify the input, de-fuzzify the output, and execute the applicable rules
d) Execute the applicable rules, fuzzify the input, and fuzzify the output

Explanation: The implementation of fuzzy logic generally involves three steps namely fuzzify the input into fuzzy membership functions, execute every rule in the defined rule base to generate the fuzzy output, de-fuzzify the output to get distinct output values.

4. What is the term used to refer to variables that take non-numerical as input?
a) String variables
b) Character variables
c) Linguistic variables
d) Morphological variables

Explanation: Linguistic variables are assigned non-numerical values to aid the representation of facts and rules. For example, a linguistic variable may take “hot” and “cold” as input. In order to take advantage of the scale offered by fuzzy logic, additional words such as “rather” or “somewhat” may be used.

5. Which among the following uses past solutions to solve current problems?
a) Artificial neural networks
b) Fuzzy logic
c) Knowledge based systems
d) Case based reasoning

Explanation: Case based reasoning is the process of developing solutions to current problems based on solutions that have already been applied under similar conditions. The process initially involves the comparison of functional and operational characteristics for resemblance. If the similarity is limited, then reasoning is applied for modification of solution to suit the current problem.

6. Which among the following is an algorithm useful for case based reasoning?
a) Viterbi algorithm
b) K- nearest neighbours algorithm
c) Forward algorithm
d) Backward algorithm

Explanation: K-nearest neighbours algorithm uses k number of training examples that have input characteristics similar to that of current condition under interest to fix target value for the current condition. This may be accomplished by using mode, average, or some other interpolation technique.

7. What is the process of identifying a parameter value from a training set that could result in the minimisation of error called?
a) K-nearest neighbours algorithm
b) Leave one out cross validation algorithm
c) Bayesian algorithm
d) Markov theorem

Explanation: The level of importance exhibited towards a parameter can be varied by adjusting the weights associated with each parameter. The degree of importance may be increased by increasing the value on scale. Leave one out cross validation algorithm is employed to learn the weights from the instances defined in the training set.

8. Case based reasoning is a cyclic process.
a) True
b) False

Explanation: Case based reasoning begins with the identification of a case that best suits the case under investigation. The case is then modified to suit the current situation. The modified case is evaluated based on performance. Finally, the new case is added to the case base according to the performance level.

9. Which among the following is used for quality determination in evolutionary algorithm?
a) Subjective function
b) Fitness function
c) Loss function
d) Utility function

Explanation: Fitness solution determines the degree of closeness exhibited by the designed solution to the specified expectations as a single figure of merit. Evolutionary algorithm determines optimal solutions by eliminating sub-optimal solutions based on the figure of merit assigned by fitness solutions.

10. Which among the following provide solutions as a string of numbers?
a) Genetic algorithm
b) Genetic programming
c) Evolutionary programming
d) Gene expression programming 