Neural Network Two Mark Q.B

NOORUL ISLAM COLLEGE OF ENGINEERING KUMARACOIL. Department of Electrical and Electronics Engineering Question Answer Bank

IC 1403 –NEURAL NETWORKS AND FUZZY LOGIC CONTROL (S8 EEE) Prepared by S.Anitha Janet Mary Lecturer/ EEE

Signature of HOD/EEE

Signature of Principal

NOORUL ISLAM COLLEGE OF ENGINEERING, KUMARACOIL. IC 1403 –NEURAL NETWORKS AND FUZZY LOGIC CONTROL 2 MARK QUESTIONS AND ANSWERS Unit 1 1. Define artificial neural network (ANN) Artificial Neural Network is information processing devices with the capability of performing computations similar to human brain or biological neural network. 2. List out the differences between artificial neural network and biological network Artificial Neural Network: i. Speed: Slow in processing information. (Processing time in the range of nanoseconds. ii. Processing: Sequential or step by step processing. iii. Size and compatibility: Simple but cannot be used for complex pattern recognition Biological network: i. Speed: faster in processing information. (Processing time in the range of milliseconds) ii. Processing: Parallel processing. iii. Size and compatibility: It can be used for complex pattern. 3. Define weight. Weight is information used by the neural net to solve a problem. 4. Define Activation Function. The activation function is used to calculate the output response of a neuron. 5. What are the classifications of activation function? 1. Identity function 2. Binary Step function 3. Sigmoidal function 6. What are the types of Sigmoidal Function? 1. Binary Sigmoidal Function 2. Bipolar Sigmoidal Function 7. What are the applications of neural networks? Used in medical field Used in telephone communication Business applications

8. Define bias. Bias acts exactly as a weight on a connection from a unit whose activation is always one. 9. What is the function of Synaptic gap? Synaptic gap is used to convert the electrical signals to some chemicals and these chemicals are again converted to electrical signals. 10. Define threshold. The threshold ‘θ’ is a factor which is used in calculating the activations of the given net. 11. What are Dendrites? Dendrites are used to receive signals from other neurons. 12. Define Training. The process of modifying the weights in the connections between network layers with the objective of obtaining the expected output is called training. 13. What are the different types of training? 1. Supervised training 2. Unsupervised training 3. Reinforcement training 14. Define Learning. Learning is the process by which the free parameters of a neural network get adapted through a process of stimulation by the environment in which the network is embedded. 15. What are the different types of learning rules? 1. Hebbian Learning rule 2. Perceptron Learning rule 3. Delta Learning rule 4. Competitive Learning rule 5. Outstar Learning rule 6. Boltzman Learning rule 7. Memory Learning rule 16. Define Back Propagation Network (BPN). It is a multi-layer forward network used extend gradient-descent waste delta learning rule.

17. What are merits and demerits of Back Propagation Algorithm? Merits: 1. The mathematical formula present here can be applied to any network and does not require any special mention of the features of the function to be learnt. 2. The computing time is reduced if the weights chosen are small at the beginning. Demerits: 1. The number of learning steps may be high, and also the learning phase has intensive calculations. 2. The training may cause temporal instability to the system. 18. What are the applications of back propagation algorithm? 1. Optical character recognition 2. Image compression 3. Data compression 4. Control problems 19. What are the four main steps in back propagation algorithm? 1. Initialization of weights 2. Feed forward function 3. Back propagation 4. Termination 20. Define supervised training. It is the process of providing the network with a series of sample inputs and comparing the output with the expected responses. 21. What is unsupervised learning rule? No teacher is present during learning process. No desired pattern is present. 22. What is Reinforcement rule? What are its types? It is output based. Teacher is present during learning process. The correctness of answer is checked and there will be no desired o/p. Rewardity is given to correct answer and penalty is given to wrong answer. Correlation and competitive rules are types of Reinforcement rule. 23. What are the types of supervised learning rule? Gradient descent rule and stochastic rule are the types of supervised rule. 24. What is Hebbian rule? Hebbian rule is reinforcement rule and is given as, Wi +1 = wi + C f(wt xi )xi Where, i = 1, 2, 3… n

25. What is perceptron rule? Wi +1 = wi + C (di – f (wt xi) xi) i = 1, 2, 3… n C is learning parameter di is desired output f(wt x) = sgn(wt x) 26. What is delta learning rule? Wi +1 = wi + C (di – f(wt xi ) f1(wt xi)xi ) 27. What is ADALINE? ADALINE is Adaptive Linear Neural Element. 28. What is Widrow – Hoff rule? Wi +1 = wi + C (di – f(wt xi )xi ) f(wt xi ) = wt xi UNIT-2 1. Define feedback networks? The networks, which can return back the output to the input, thereby giving rise to an iteration process, are defined as feedback networks. 2. Give some examples of feedback networks Some examples of feedback networks are Simulated annealing, Boltzmann machine, Hop field net, etc. 3. What is the main purpose of Hop field network? Hop field network is able to recognize unclear pictures correctly. However, only one picture can be stored at a time. 4. Define discrete Hop field net The discrete Hop field net is a fully interconnected neural net with each unit connected to every other unit. The net has symmetric weights with no selfconnections i.e. all the diagonal elements of the weight matrix of a Hop field net are zero 5. List two difference between Hop field and iterative auto associative net. The two mail differences between Hop field and iterative auto associative net are that, in the Hop field net, Only one unit updates its activation at a time, and, Each unit continues to receive an external signal in addition to the signal from the other units in the net.

6. What is energy function or Lyapunov function? The asynchronous discrete time updating of the units allows a function known as energy function or Lyapunov function. This function proves that the net will converge to a stable set of activations. 7. Write the energy function for discrete Hop field network. The energy function for the discrete Hop field network is given by E= -0.5_(ij) _(j) yi yj wij - _ xi yi + _ (i) i yi. 8. What is continuous Hop field net? The continuous Hop field net is a modified form of discrete Hop field net. It uses continuous valued output functions, which can be used either for associative memory problems or constrained optimization problems. 9. Write the energy function of continuous Hop field network. The energy function continuous Hop field network is, E = 0.5 _(i=1 to m) _(j=1 to m) wij vij vj + _(i=1 to m) i vi . 10. What are the basic building blocks of artificial neural network? The basic building blocks of the artificial neural network are: 1. Network architecture. 2. Setting the weights. 3. Activation function. 11. State the purpose of process identification The purpose of process identification is to find the response y of the plant for the given set of inputs x. 12. List the types of plant identification. A) Forward plant identification. B) Plant reverse identification . 13. State some disadvantages of forward plant identification Forward plant identification is always feasible; however, it does not immediately allow for construction of the plant controller. 14. What is the goal of inverted pendulum? The goal of inverted pendulum task is to apply a sequence of right and left forces of fixed magnitude such that the pendulum is balanced, and the cart does not hit the edge of the track. 15. State the property of inverted pendulum neurocontroller. This technique is the simple and synergistic approach, which examines the merger of visual image acquisition, training, and neurocontrol.

16. What are the variables, which describe the state of the system? Four variables describe the state of the system: the horizontal position and the velocity of the cart (x, v), the angle between the pendulum and vertical, and the angular velocity of the falling pendulum (, ). 17. State the force required to stabilize the system to be controlled in inverted pendulum. The force required to stabilize the system is a function of four state variables x, v, , and as follows, F(t) = ax(t) + bdx(t)/dt + c(t) - d(t). Where, dx (t)/dt = x (t), d(t)= (t) and a, b c and d are constant coefficients. 18. Define neural network The neural network can also be defined as an inter connection of neurons, such that neuron outputs are connected, through weights, to all other neurons including themselves; both lag-free and delay connections are allowed. 19. State the need for training the neural network The observer who would measure the cart position and angles of pendulum at every instant is replaced by the input of the visual image of the cart and the pendulum. Since it is essential to provide the controller with the present and most recent position and angle data, the present and most recent images arranged in pairs have been used to train the network. 20. How ANN resembles brain? The ANN resembles the brain in two respects: 1) Knowledge is acquired by the network through a learning process, and, 2) Inter-neuron connection strengths known as synaptic weights are used to store the knowledge. 21. State how a feedback network can be obtained from a feed forward network? A feed back network can be obtained from a feed forward network by connecting the neuron’s outputs to their inputs. UNIT -3 1. What are classical sets? Classical sets are objects that satisfy precise properties of membership. It has a crisp set boundary 2. List the operations on classical sets? Union AUB= {x |x € A or x € B} Intersection A_B= {x |x € A and x € B} Complement A= {x | x € A, x € X}

Difference A | B= {x |x € A and x € B} 3. List the properties of crisp sets? Commutativity AUB=BUA A∩B= B∩A Associativity AU (BUC) = (AUB) UC A∩ (B∩C) = (A∩B) ∩C Distributivity AU (B∩C) = (AUB) ∩ (AUC) A∩ (BUC) = (A∩B) U (A∩C) Idempotency AUA =A A∩A=A Identity AUØ=A A∩X=A A∩Ø=Ø AUX=X Transitivity If A с B с C, then A с C Involution A=A 4. State the excluded middle laws and De Morgan’s laws for classical sets. Excluded Middle laws: Law of excluded middle AUA=X Law of contradiction A∩ A=Ø De Morgan’s Laws: (A∩B) c = A c U B c (A U B) c = A c ∩ B c 5. What are Fuzzy sets? Sets containing elements that have varying degrees of membership in the set are called fuzzy sets. 6. List the Fuzzy set operations? Union μAUB(x) = μA(x) μB(x) Intersection μA_B(x) = μA(x) μB(x) Complement μA (x) =1-μA(x) 7. List the properties of fuzzy sets? Commutativity AUB=BUA A∩B= B∩A Associativity AU (BUC) = (AUB) UC A∩ (B∩C) = (A∩B) ∩C Distributivity AU (B∩C) = (AUB) ∩ (AUC) A∩ (BUC) = (A∩B) U (A∩C) Idempotency AUA =A A∩A=A Identity AUØ=A A∩X=A

A∩Ø=Ø AUX=X Transitivity If A с B с C, then A с C Involution A=A 8. Give De Morgan’s law and Excluded middle laws for fuzzy sets? De Morgan’s Laws: (A∩B) c = A c U B c (A U B) c = A c ∩ B c Excluded middle laws: AUA c ≠ X A∩A c ≠Ø 9. What you mean by Universal set? Set of all elements in the universe is termed as Universal set X={x1, x2, x3…….} 10. Define Cardinality number. Number of elements in the universe of discourse is called Cardinality number. It is denoted by nx 11. Give the expression for cardinality of power set. P=2nx Cardinality of power set is infinity 12. Define fuzzification. Fuzzification is the process of converting a crisp quantity into a fuzzy quantity. 13. What are fuzzy relations? Fuzzy relations R is a mapping from Cartesian space X ×Y to the unit interval [0,1] where the strength of mapping is expressed by the membership function of the relation for ordered pairs from the two universal or μR(x,y) 14. List the operations on fuzzy relations. Union μRUS (x,y)=max(μR(x,y), μS(x,y)) Intersection μR_S (x,y)=min(μR(x,y), μS(x,y)) Complement μR(x,y)= 1- μR(x,y) Containment R S=> μR(x,y) _μS(x,y) 15. List the properties of fuzzy relations. RU R = E R∩R = Ø

16. Define Defuzzification. Defuzzification is the conversion of a fuzzy quantity into crisp quantity. 17. Differentiate fuzzification and defuzzification? Fuzzification It is the conversion of crisp quantity into fuzzy quantity. Defuzzification It is the conversion of a fuzzy quantity into crisp quantity. 18. List the defuzzification methods. _ Max – membership principle _ Centroid methods _ Weighted average methods _ Mean-max membership _ Center of sums _ Center of largest area _ First(or last) of maxima 19. Differentiate Classical set Fuzzy set Classical set They have objects that satisfy precise properties of membership. Fuzzy set have the objects that satisfy imprecise properties of membership. 20. Explain the defuzzification method of center of sums This process involves the algebraic sum of individual output fuzzy sets instead of their union. Here the weights are the areas the respective membership functions. _z z _n k=1μCk (z) d z Z* =_z _n k=1μCk (z) d z 21. What is Cartesian product? Cartesian product of two crisp sets X & Y is denoted by X × Y is a set of all ordered pairs.

UNIT –IV 1. What are the features of membership functions?  Core  Support  Boundary

2. Define core of a membership function. The core of a membership function is defined as the region of universe that is characterized by complete and full membership function in the set. 3. Define support of a membership function. The support of a membership function is defined as the region of universe that is characterized by non-zero membership function in the set. 4. Define boundaries of a membership function. The support of a membership function is defined as the region of universe that is characterized by non-zero membership function but not complete membership in the set. 5. What is a normal fuzzy set? A normal fuzzy set is one whose membership function has atleast one element x in the universe whose membership value is unity. 6. Define prototype of the set. For fuzzy sets where one and only one element has a membership function equal to one, this element is referred to as prototype of the set. 7. Define a convex fuzzy set. A convex fuzzy set is described by the membership function whose membership values are monotonically increasing, or decreasing. 8. Define cross over points of a membership function. The support of a membership function is defined as the elements of the universe for which a particular fuzzy set has values equal to 0.5. 9. Define height of a fuzzy set. Height of a fuzzy set is the maximum value of membership function in a given set. 10. List the three operators in GA Cross over, Mutation and Reproduction. 11. Define Reproduction. Reproduction is the process by which strings with better fitness values receive correspondingly better copies in the new generation. 12. Define Mutation. Mutation is the process in which strings are able to mix and match their desirable qualities in a random fashion. 13. Define crisp ordering. For issues or actions that are deterministic there is no ambiguity in the

ranking ,we call this as crisp ordering. 14. Define degree of consensus. The individual preference of those in the decision group are collected to form a group metric whose properties are used to produce a scalar measure of “degree of consensus”. 15. What are the two common measures of preference? The two common measures of preference average fuzziness and average certainity. 16. What is Classical Bayesian decision method? These presume that future states of nature can be characterized as probability events. 17. What is meant by Rank ordering? Assessing preference by a single individual, a committee, a poll and other opinion methods can be used to assign membership values of a fuzzy variables. 18. What are the types of searching techniques in Genetic Algorithm? 1. Calculus based 2. Guided random 3. Enumerative 19. What are the types of Genetic Algorithm? 1. Parallel 2. Sequential 20. What are the methods involved in Genetic Algorithm? 1. Simulated annealing 2. Ant colony optimization 3. Random cost strategy 5. Cellular automata 21. What is the Ant colony optimization? The foraging of ants has led to novel algorithm called Ant colony optimization for rerouting network traffic in busy telecommunication systems. 22. Mention the three important aspects of Genetic Algorithm? 1. Definition of objective function 2. Definition and implementation of genetic representation. 3. Definition and implementation of genetic operators. 4. Evolution

UNIT –V 1. What are the basic elements of a fuzzy logic control system? Scaling factors,Fuzzification,Defuzzification,Knowledge base,Decision making Logic. 2. Give the structure of a fuzzy production rule system. A set of rules that represents the policies and heuristic strategies of expert decision maker. 3. What are the assumptions to be made in a fuzzy control system design? i) The plant is observable and controllable. ii) A solution exists. 4. Define approximate reasoning The input data, rules and output action, or consequences are generally fuzzy sets expressed as membership functions defined in a proper space. The method used for evaluation is approximate reasoning. 5. What is the purpose of Knowledge base module? It contains knowledge about the input and output fuzzy partitions. It will include the term set and the corresponding membership functions defining the input variables to fuzzy rule based systems. 6. Explain the steps in designing a fuzzy control system. i) Identify the variables of the plant. ii) Partition the universe of discourse into a number of fuzzy subsets. 7. List the features of fuzzy control system. i) Fixed and uniform input-output scaling factors. ii) Fixed membership functions. 8. Give the differential equation in Inverted pendulum. -ml2 d2 _/dt2 + mlg sin _ = u(t) 9. What is the automating the control of depth of anesthesia. It is to release the anaesthtist so that he or she can devote attention to other tasks,such as controlling fluid balance. 10. Why modeling of the process, blood pressure control difficult. Because the biological process like anaesthesia has a non-linear time varying strcture,it suggests the use of rule based controllers. 11. How the depth of anesthesia is controlled. The depth of anesthesia is controlled by a mixture of drugs that are injected intravenously or inhaled as gases.

12. What are the gases inhaled during anaesthesia. Isoflourane is widely used, more often in a mixture of 0 to 2 percent by volume of isoflorane in oxygen or nitrous oxide. 13. What are the two different types of disturbances? 1. Surgical disturbances 2. Measurement noise and artifacts. 14. What is the purpose of a large sensor image chip? It detects the motion vector and a fuzzy system decides if the motion is due to trembling. 15. What is the purpose of fuzzy control system in a TV? It controls the contrast, brightness, velocity modulation and sharpness. 16. Define an adaptive fuzzy system. The input values are normalized and converted into fuzzy representations; the models rule base is executed in parallel fashion to produce a consequent fuzzy region. 17. List some of the applications of fuzzy logic control system. Inverted pendulum, Home heating system , Blood pressure during Anesthesia. 18. What are the two types of control systems? 1. Open loop control system 2. Closed loop control system 19. What is meant by a sensor? The system for measurement of the controlled signal is called a sensor. 20. Define plant. The physical system under control is called a plant. 21 State control problems. The control problem states that the output or response of the physical system under control is adjusted as required by the error signal. The error signal is the difference between the actual response of the plant as measured by the sensor system and the desired response. As specified by a reference input. 22. Define control surface. The control surface describes the dynamics of the controller and is generally a time-varying non-linear surface. 23. Mention the fuzzy logic controller.  Fuzzification strategies.

   

Knowledge base. Rule – base. Decision making logic. Defuzzification strategies. 16 MARK QUESTIONS UNIT-I

1. Give artificial neural network architectures? - Write the types of architectures - Draw the architectures diagrams - Write the formulas. 2. Explain Single layer feed forward network - Draw the network diagram - Write the step by step procedure. 3. Explain Multilayer feed forward network - Draw the network diagram - Write the step by step procedure. 4. Explain the Hebbian rule? - Derive the learning algorithm - Draw the network diagram 5. What is perceptron rule and briefly explain it. - Derive the learning algorithm - Draw the network diagram 6. Explain the various types of learning paradigms in an ANN. -Supervised learning algorithm - Unsupervised learning algorithm - Reinforcement learning algorithm 7. What are the factors to be considered for employing artificial neural networks? - Choice of model - Learning algorithm - Robustness 8. Explain some types of neural networks. - Feed forward neural network - Recurrent network - Hopfield network

9. Give some applications of ANN    

Function approximation, or regression analysis, including time series prediction, fitness approximation and modeling. Classification, including pattern and sequence recognition, novelty detection and sequential decision making. Data processing, including filtering, clustering, blind source separation and compression. Robotics, including directing manipulators, Computer numerical control. .

10. Explain about McCullough-Pitts neuron. 1. 2. 3. 4. 5.

A set of n excitatory inputs, xi. A set of m inhibitory inputs, xn+j A threshold, u. A unit step activation function. A single neuron output, y. UNIT-II

1) Distinguish between hop field continuous and discrete models. - Draw the network diagrams - Write the differences of the two networks 2) Explain briefly the back propagation technique. - Write the step by step procedure for learning algorithm - Draw the architecture. 3) Explain how the ANN can be used for process identification with neat sketch. - Types of PI methods - Diagrams of plants - Derive the plant equation 4) Discuss the step by step procedure of back propagation learning algorithm in detail. - Write the ten steps in BPN 5) Explain the transient response of continuous time networks. - Draw the network - Write the state equation - Apply the continuous hop field algorithm - Derive the transient response 6) Explain how ANN can be used for neuro controller for inverted pendulum. - Draw the inverted pendulum - Write the input output parameters

- Write the state equation - Simulation results UNIT-III 1) Differentiate fuzzy set from classical set and name the properties of classical (crisp) sets. Commutativity AUB=BUA A∩B= B∩A Associativity AU (BUC) = (AUB) UC A∩ (B∩C) = (A∩B) ∩C Distributivity AU (B∩C) = (AUB) ∩ (AUC) A∩ (BUC) = (A∩B) U (A∩C) Idempotency AUA =A A∩A=A Identity AUØ=A A∩X=A A∩Ø=Ø AUX=X Transitivity If A с B с C, then A с C Involution A=A 2) A = {(1/2) + (0.5/3) + (0.3/4) + (0.2/5)}, B = {(0.5/2) + (0.7/3) + (0.2/4) + (0.4/5)} Calculate the several operation of the fuzzy set. -Union μAUB(x) = μA(x) μB(x) -Intersection μA_B(x) = μA(x) μB(x) -Complement μA (x) =1-μA(x) 3) Discuss varies properties and operations on crisp relation. - Cartesian product - Max-min composition 4) Describe fuzzy relation. - Cartesian product - Max-min composition 5) Explain the operation of fuzzy sets with a suitable example. Union μAUB(x) = μA(x) μB(x) Intersection μA_B(x) = μA(x) μB(x) Complement μA (x) =1-μA(x) 6) Define defuzzification and explain the different defuzzification methods. _ Max – membership principle _ Centroid methods _ Weighted average methods

_ Mean-max membership _ Center of sums _ Center of largest area _ First (or last) of maxima

UNIT-IV 1) Write the components of a fuzzy logic system and explain them.  Fuzzification strategies.  Knowledge base.  Rule – base.  Decision making logic.  Defuzzification strategies. 2) Explain min-max method of implication with a suitable example. - Find the Cartesian product of the relations - Apply the min max composition 3) Explain the various ways by which membership values can be assigned to fuzzy variables.  It is a curve that defines how each point in the input space is mapped to a membership value between 0 and 1.  Membership function in a mapping of X on [0,1]  It gives the degree of membership of a element in a given fuzzy set. 4) Discuss the various special features of the membership function.  Core  Support  Boundary 5) With a neat sketch discuss the major components of fuzzy controller. - Fuzzy rule – based structures - Fuzzy relational equations 6) Write about genetic algorithm and its application. - Cross over - Mutation - Reproduction. UNIT-V 1) Explain the importance of fuzzy logic control in various fields. - Control applications -image processing

- Home heating system - Medical applications 2) Explain the fuzzy logic is being implemented for image processing.      

Document the system's operational specifications and inputs and outputs. Document the fuzzy sets for the inputs. Document the rule set. Determine the defuzzification method. Run through test suite to validate system, adjust details as required. Complete document and release to production.

3) Discuss the home heating system with fuzzy logic control. - draw the diagram - explain the FLC - Explain the parts - Explain the controlling process 4) Explain the technique “fuzzy logic blood pressure during anesthesia” in a brief manner. -Write the disturbances -draw the controlling diagram -Derive the control equation 5) What are the components of fuzzy logic control and explain them in detail with block diagram?  The plant is observable and controllable. State, input and output variables are usually available for observation and measurement or computation.  The control engineer is looking for a “good enough” solution, not necessarily the optimum one.