CS607_Artificail Intelligence_Past_FinalTerm_Papers_Vuhelp.pk
My today paper 22/08/2014 08:00am
Total marks 80
Mcqs = 40
5 mrks Q = 4
3 mrks Q = 4
2 marks Q = 4
(5 marks Questions)
- Application areas of fuzzy system. (page 153)
- Elaborate Action predicates in SRTIPS. (page198
- Elaborate ID3 algorithm (page177
- General stages of expert system (page129
(3 marks Questions)
- Which sub-field of AI is concerned about enabling machine to see, perceive and understand? Briefly explain. (page203
- Briefly explain Law of Excluded Middle. (page145
- Main phase of linear sequence. (page129
- Which is best term suited for situation in which data in a problem can be separated into their respective classes by using single straight line. Also give example. (page184, 185
(2 marks Questions)
- What do you know about the term STRIPS. (page197
- Elaborate the term Genetic Algorithm. (page77
- Which topic in AI deal with the concept of partial truth and member function. (page147
- Out of thetrainingandvalidation, which term is best suited for situation in which we give different training examples of correct behavior to machine and then train that machine on these training examples in order to make that machine to learn.
Best of luck
Plz post ur current paper here
And remember me in ur prayers.
cs607
FIS(Fuzzy inference system) kahan kahan use hoti ha (page153
Computer and insan main kn best ha table bna k likhna tha (page182
Unsupervised learning brief (page190
Agr instance space increase ho gi to kia concept space increase ho gi? (page167
AI ki wo kn si field ha jo computer ko insan jesa sochny jesa bnati ha (page203
Clips main father n son ka relationship show and assert kraen (page134
Deductive learning with example (page162
Group with out labels kia hota ha
Mid term k mcqs prh k jaen
AI main kia kia 0 to 1 k drmian ha which terms u have studied (ans probability, Entropy, etc
Aisi problems jo single line se separate na ho .... unhaen kn separate nai kr skta ik example b daen
Degree of truth of fuzy logic kes sy show hota (page147
finalterm paper cs607 23 august
compare biological neural network with computer detail ...5(page182
class set is container which doesnot wholy includ or exclude ...agree or not justify...(page145
what is machine learing ...3
wht u study about computer vision....5(page203
clips stand for ....mcq(page133
fuzzy has nature ....mcq
mine todays paper..These are some questions which I remember till now..
what is concept space and instance space? (page166, 167
describe fuzzy logic and crisp function in terms of membership. (page147
briefly describe law of excluded middle (page145
what do you study in computer vision? (page203
one question was of CLIPS already shared by Asif (MCS 4th). (page133
applications of fuzzy logic? (page147
briefly describe action predicate? (page198
two main types of problems and their subtypes. (page165
My today papers 26-08-14
1. How Machine learning help in developing expert systems (5marks) (page163
2. Action Predicates in STRIPS with example (5 marks) (page198
3. Five parts of the fuzzy inference process (5 marks) (page154
4. Write name of five applications in which we apply learning algorithms. (5 marks) (page163
5. Name of design phase in which ANNs start learning (3marks) (page188
6. Advantages of ANNs (3marks) (page187
7. Defuzzification --> Rule evaluation ---> Fuzzification is a sequence of steps to design fuzy logic machine. Are you agree with the statement also give correct statement (3marks) (page154
8. Names of emergent fields in which learning algorithms are used. (3 marks) (page163
9. Which step of FIS resolve all antecedent statement to a degree of membership between 0 and 1. (2 marks) (page147
10. Knowledge elicitation and its methods (2 marks) (page130
11 Clustering defination (2 marks) (page205
12 Information gain defination (2 marks) (page177
According to my point of view Question no 3 and Question no 7 have same answer which is
• Fuzzification of the input variables
• Application of fuzzy operator in the antecedent (premises)
• Implication from antecedent to consequent
• Aggregation of consequents across the rules
• Defuzzification of output
Similarly Question no 4 and Question no 8 have same answer which is
• Spoken digits and word recognition
• Handwriting recognition
• Driving autonomous vehicles
• Path finders
• Intelligent homes
• Intrusion detectors
• Intelligent refrigerators, tvs, vacuum cleaners
• Computer games
• Humanoid robotics
But please verify it
Totally from Moaaz file
MCQS half were new and half from past papers
CS607 Final term Paper August 22, 2014
Total 1 Mark MCQ = 40
Total 2 Marks Short Questions = 4
Total 3 Marks Short Questions = 4
Total 5 Marks Long Questions = 4
Some Objective
A single layer perceptron cannot perform pattern classification on linearly separable patterns.
►True
►False (Page 186)
______is the process by which the fuzzy sets that represent the outputs of each rule are combined into a
single fuzzy set.
►Aggregation (Page 157)
►Fuzzification
►Implication
►None of the given
The tractable problems are further divided into structured and ______problems
►Non-structured
►Complex (Page 166)
►Simple
A proposition is the statement of a ______.
►Fact (Page 98)
►Equation
►Action
►Theorem
The Candidate-Elimination algorithm represents the ______
►Version space (Page 173)
►Solution space
►Elimination space
►None of the given
Identify the step involved in planning phase.
1. Knowledge acquisition from expert
2. Coding
3. Resource allocation (Page 129)
4. Identify concrete knowledge element
MCq’s where easy mostly from moaaz file and some were from other files too
Subjective
1)What is the law of ‘Excluded Middle’? (2)
Answer:-
It was Aristotle who came up with the ‘Law ofthe Excluded Middle’, which states that any element X, must be either in set A or in set not-A. It cannot be in both. And these two sets, set A and set not-A should contain the entire universe between them.
2)What is state? (2)
Answer:
State is a conjunction of predicates represented inwell-known form, for example,
a state where we are at the hotel and do not have either cash or radio is
represented as,
at(hotel) ∧ ¬have(cash) ∧ ¬have(radio)
3)Design phases of ANNs?? (5)
Answer:
Design phases of ANNs
Feature Representation
The number of features are determined using no of inputs for the Problem. These large feature spaces make algorithms run slower. They also make the training process longer. The solution lies in finding a smaller feature space which is the subset of existing features.
Feature Space should show discrimination between classes of the data. Patient’s height is not a useful feature for classifying whether he is sick or healthy .
Training
Training is either supervised or unsupervised.
Remember when we said:
We assume that the concept lies in the hypothesis space. So we search for a hypothesis belonging to this hypothesis space that best fits the training examples, such that the
output given by the hypothesis is same as the true output of concept
Finding the right hypothesis is the goal of the training session. So neural networks are doing function approximation, and training stops when it has found the
closest possible function that gives the minimum error on all the instances
Training is the heart of learning, in which finding the best hypothesis that covers most of the examples is the objective. Learning is simply done through adjusting the weights of the network
Similarity Measurement
A measure to tell the difference between the actual output of the network while training and the desired labeled output The most common technique for measuring the total error in each iteration of the neural network (epoch) is Mean Squared Error (MSE).
Validation
During training, training data is divided into k data sets; k-1 sets are used for training, and the remaining data set is used for cross validation. This ensures better results, and avoids over-fitting.
Stopping Criteria
Done through MSE. We define a low threshold usually 0.01, which if reached stops the training data. Another stopping criterion is the number of epochs, which defines how many maximum times the data can be presented to the network for learning.
Application Testing
A network is said to generalize well when the input-output relationship computed by the network is correct (or nearly so) for input-output pattern (test data) never used in creating and training the network.
4)What are action predicates? Also give some examples (5)
Answer:
Action is a predicate used to change states. It has three components namely, the
predicate itself, the pre-condition, and post-condition predicates. For example,
the action to buy something item can be represented as,
Action:
buy(X)
Pre-conditions:
at(Place) ∧sells(Place, X)
Post-conditions/Effect:
have(X)
What this example action says is that to buy any item ‘X’, you have to be (preconditions) at a place ‘Place’ where ‘X’ is sold. And when you apply this operator
i.e. buy ‘X’, then the consequence would be that you have item ‘X’ (postconditions).
5)Application of Fuzzy inference system (5)
Answer:- Page :153
Fuzzy inference systems have been successfully applied in fields such as
automatic control, data classification, decision analysis, expert systems, and
computer vision.
6)What is FIS system ? (5)
Answer:- Page :153
Fuzzy inference system (FIS) is the process of formulating the mapping from a given input to an output using fuzzy logic. This mapping then provides a basis from which decisions can be made, or patterns discerned Fuzzy inference systems have been successfully applied in fields such as automatic control, data classification, decision analysis, expert systems, and
computer vision. Because of its multidisciplinary nature, fuzzy inference systems are associated with a number of names, such as fuzzy-rule-based systems, fuzzy expert systems, fuzzy modeling, fuzzy associative memory, fuzzy logic controllers, and simply (and ambiguously !!) fuzzy systems. Since the terms used to describe the various parts of the fuzzy inference process are far from standard, we will try to be as clear as possible about the different terms introduced in this section.
7)Find out the value of following logical operators using Fuzzy-And. (3)
Answer:
A table of values was given like.
Read the following link to find out more about min(A and B) and max(A OR B).
A / B / A and-fuzzy B0.7 / 0.3 / 0.3
0 / 0.4 / 0
1 / 1.2 / 1
8)Which term is best suited for the situation where we have combination of disjunction and conjunction? (3)
Answer: Page : 176
Decision trees give us disjunctions of conjunctions, that is, they have the form:
(A AND B) OR (C AND D)
In tree representation, this would translate into:
9)What is ID3 and how it works?? (3)
Answer: Page : 177
ID stands for interactive dichotomizer. This was the 3rdrevision of the algorithm
which got wide acclaims. The first step of ID3 is to find the root node. It uses a special function GAIN, to evaluate the gain information of each attribute. For example if there are 3 instances, it will calculate the gain information for each. Whichever attribute has the maximum gain information, becomes the root node. The rest of the attributes then fight for the next slots.
10)At which phase in ANNs design system starts learning? (3)
Answer:
Training
Training is either supervised or unsupervised.
Remember when we said:
We assume that the concept lies in the hypothesis space. So we search for a hypothesis belonging to this hypothesis space that best fits the training examples, such that the
output given by the hypothesis is same as the true output of concept
Finding the right hypothesis is the goal of the training session. So neural networks are doing function approximation, and training stops when it has found the
closest possible function that gives the minimum error on all the instances
Training is the heart of learning, in which finding the best hypothesis that covers most of the examples is the objective. Learning is simply done through adjusting the weights of the network
11)What are the steps of linear model of ES? (2)
Answer: Page: 129
A linear sequence of steps is applied repeatedly in an iterative fashion to develop the ES. The main phases of the linear sequence are
• Planning
• Knowledge acquisition and analysis
• Knowledge design
• Code
• Knowledge verification
• System evaluation
12)Single layer perception (2)
Answer: page 184
A single layer perception can perform pattern classification only on linearly
separable patterns, regardless of the type of non-linearity (hard limiter, signoidal).
itnaaaaaaaaa mushkil, so conceptual
puri book ko chaan chaan k mer gai but phir b subjective nai ata tha, pta hi nai chala k kia or kidher sy aya
well, jo yad ha wo post ker rhi hun bas dua karna k papers achy ho jaen.
Write down the names of some areas where clustering is benifitial.
How mechanical intelligence impact the Robot.
how ANNs manages their w8s.
Disjunctive space
FindS algorithm
compare different searches each other and which is best.
define domain of problem.
Version space and ID3 are connected with each other, agree or not, give solid reasons.
salam to all
MCQS 40
4 QUIZ OF 2 MARKS
4 QUIZ OF 3 MARKS
4 QUIZ OF 5 MARKS
objective mostly past papers but statements change.
or subjective last walya 3 chapters sa tha.
- about machine learning
- NEURAL NETWORK k 3 sa 4 quiz thya or mcqs b thya.
- 3 sa 4 quiz ma senrio given tha or batna tha kon sa alogrithme batter h aink liya.
- quiz about Rules
- quiz about ESDLC
- quiz about learning agent or algoritm
MCQZ k liya mozz file dakh lain .
TOTAL MCQS 40
35 MCQS WAS MOAZ FILE
- un-supervised learning
- inductive learning
- write the name of predicate action?
- what is STRIPS?
- comparison b/t brain and computer?
- Elaborate the five parts of fuzzy infrence system?
- Write the main phases of linear sequance of ESDLC?
- Patteren recognition in computer vision?
- backword chain steps define?
un-supervised learning
Clustering is a form of unsupervised learning, in which the training data is
available but without the classification information or class labels. The task of
clustering is to identify and group similar individual data elements based on some
measure of similarity.
write the name of predicate action?
the predicate itself, the pre-condition, and post-condition predicates.what is STRIPS?
STRIP is one of the founding languages developed particularly for planning. Let us understand planning to a better level by seeing what a planning language can represent.comparison b/t brain and computer?
Elaborate the five parts of fuzzy inference system?
Five parts of the fuzzy inference process
•Fuzzification of the input variables
•Application of fuzzy operator in the antecedent (premises)
•Implication from antecedent to consequent
•Aggregation of consequents across the rules
•Defuzzification of output
Write the main phases of linear sequance of ESDLC?
The main phases of the
linear sequence are
Planning
Knowledge acquisition and analysis
Knowledge design
Code
Knowledge verification
System evaluation
Pattern recognition in computer vision?
Computer vision encompases topics from pattern recognition, machine learning,
geometry, image processing, artificial intelligence, linear algebra and other
subjects.
backword chain steps define?
1. Start with the goal.
2. Goal may be in WM initially, so check and you are done if found!
3. If not, then search for goal in the THEN part of the rules (match
conclusions, rather than premises). This type of rule is called goal rule.
4. Check to see if the goal rule’s premises are listed in the working memory.
5. Premises not listed become sub-goals to prove.
6. Process continues in a recursive fashion until a premise is found that is
not supported by a rule, i.e. a premise is called a primitive, if it cannot be
concluded by any rule
7. When a primitive is found, ask user for information about it. Back track and
use this information to prove sub-goals and subsequently the goal.
Mcqz form clips and fuzzi systems mostly
(1)What is Goal in STRIPS given an example 2marks
Goal
Goal is also represented in the same manner as a state. For example, if the goalof a planning problem is to be at the hotel with radio, it is represented as,
at(hotel)_have(radio)
(2) elaborate the inductive learning 2mrks
Inductive learning
Inductive learning takes examples and generalizes rather than starting with existing knowledge.
(3) Boolean logic is a subset of fuzzi logic if u agree give reason?
Fuzzy logic is a superset of conventional (Boolean) logic that has been extended to handle the concept of partial truth -- truth values between "completely true" and "completely false".
(4) ESDLC steps?
•Feasibility study • Rapid prototyping • Alpha system (in-house verification) • Beta system (tested by users) • Maintenance and evolution(5) write 3 phases of ES life cycle
Genetic Algorithm is inspired by the structure and/or functional aspects of the biological neural networks and it consist of an interconnected group of artificial neurons. Do you agree or not? 5 marks(7) advantages of neural network?
•Excellent for pattern recognition•Excellent classifiers
•Handles noisy data well
•Good for generalization
Who does neural network resemble the human brain.
It resembles the brain in two respects:
Knowledge is acquired by the network through a learning process (called
training)
Interneuron connection strengths known as synaptic weights are used to
store the knowledge
(9)Elaborate version space ?
Version space is a set of all the hypotheses that are consistent with all the
training examples. When we are given a set of training examples D, it is possible
that there might be more than one hypotheses from the hypothesis space that are
consistent with all the training examples.
(10)generic algorithm inspired by the and/or structure of biological neural networks
Ifu r agree give reason 5 mrk
Elaborate the 2 main branches of space but I forget the name of space ma be hypothesis space or version space I don’t knw 5 mrks
CS604 current Final term paper 1-3-2014
MCQs half were new and half were old!
Q1) What is Goal in Strips? Give an example. (2 Marks)
Goal is also represented in the same manner as a state. For example, if the goal of a planning problem is to be at the hotel with radio, it is represented as,
at(hotel) ^ have(radio)
Q2) Write two fields or data types of CLIPS? (2 Marks)
Fields are the main types of fields/tokens that can be used with clips. They can be:
1. Numeric fields: consist of sign, value and exponent
• Float .e.g. 3.5e-10
• Integer e.g. -1 , 3
2. Symbol: ASCII characters, ends with delimiter. e.g. family
3. String: Begins and ends with double quotation marks, “Ali is Ahmed’s brother”