Scribd
Upload
21 views4 pages

Assignment2 Deadline2080 Magh 19

Uploaded by

neupaneprejan2018
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
21 views4 pages

Assignment2 Deadline2080 Magh 19

Uploaded by

neupaneprejan2018
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

BSc.

CSIT /4th Semester/ AI/ Assignment-2

Assignment- 2

Instructions:
 Handwritten answer must be submitted either in A4 sized paper or loose sheets.
 Assignment must be submitted within specified deadline and with proper cover page by clearly
mentioning subject name, assignment number , submitters information (name, roll-number and
section )
 The answers should be written after understanding the concepts and blindly copying from
others is strictly discouraged.
Recommendations:
 Students are highly recommended to follow text-books and internationally recognized
reference books, research papers and authentic websites/blogs/tutorials.

2. Solve the following problems:


2.1.Discuss about following knowledge representation techniques (with syntax/concept and
examples)
a) Frames
b) Semantic Net
c) Rule Based System (Production System)
d) Scripts
e) Conceptual Dependencies
2.2. Introduce Propositional logic and predicate logic with their strength and weaknesses.
2.3.Explain the concepts of unification and lifting in predicate logic with examples.
2.4.Convert following sentences into FOPL and prove that “Roney is naughty” using
resolution algorithm.

All over smart persons are stupid


Children of all stupid persons are naughty
Roney is Children of Harry
Harry is over smart
2.5.Explain about forward chaining and backward chaining. Differentiate between them.
2.6. Explain about following with suitable examples:
a) Inference using Full Joint Distribution
b) Bayesian Networks and Reasoning in Belief Networks
2.7.How belief networks are constructed? Consider the probability of having cloudy is 50%.
The probability that it will rain given the conditions it will be cloudy and if it is winter is

1
Collected by Bipin Timalsina
BSc.CSIT /4th Semester/ AI/ Assignment-2

30%. The probability of being winter is 50%. The probability that it will be shiny is 70%.
Now construct a belief network for this example .
2.8. Explain the concept of fuzzy logic. Describe about :
 Fuzzy Set
 Membership functions in Fuzzy Set
 Fuzzy Rule Base System
2.9.Explain the significance of Machine learning in AI.
2.10. Discuss in Detail about followings with key characteristics and challenges:
 Supervised Machine Learning
 Unsupervised Machine Learning
 Reinforcement Learning
2.11. Explain about Naive Bayes Classifier with suitable example.
2.12. Explain the concept of evolutionary algorithm and genetic algorithm. Discuss the
working of selection, crossover and mutation operations in genetic algorithm with
examples.
2.13. Explain how genetic algorithm works with proper flow diagram.
2.14. Explain the concept of neuron and neural network in AI. Discuss about mathematical
model of ANN. Briefly describe the concept of activation functions with examples.
2.15. Explain about feed forward and feed backward neural networks.
2.16. Discuss about following machine learning algorithms
a) Perceptron learning
b) Hebbian Larning
c) Back propagation Learning
2.17. Consider the following feed-forward ANN with input, weight and bias values as in
given table. (xi , wij and θi represents inputs, initial weights and initial biases
1
respectively.) If activation function is sigmoid i.e. 𝑓(𝑥) = 1+𝑒 −𝑥 calculate the outputs at
neurons 4, 5 and 6.

2
Collected by Bipin Timalsina
BSc.CSIT /4th Semester/ AI/ Assignment-2

2.18. Consider above ANN and values in (Q.NO. 2.16) and trace a single iteration of
backpropagation algorithm for updating weights and biases with training tuple X =
(1,0,1) and output label 1.
[ Assume learning rate is 0.9 and Error with output layer is computed as ( T is labeled
(targeted or exact) output and O is computed output)

and error at hidden layer is computed as :

Where l is the learning rate.

3
Collected by Bipin Timalsina
BSc.CSIT /4th Semester/ AI/ Assignment-2

]
2.19. Define expert system. Explain structure of expert system with major components.
Explain the phases in development of expert system.
2.20. Discuss about Natural language understanding and natural language generation.
Explain about steps in NLP including Lexical Analysis(Segmentation, Morphological
Analysis), Syntatic Analysis, Semantic Analysis, Pragmatic Analysis
2.21. Discuss about different application areas of NLP.
2.22. Discuss the concept of machine vision with major components of machine vision
system. Discuss about applications of machine vision.
2.23. Describe Robotics as an application of AI with Robot hardware (sensor and effectors)
and robotic perceptions.

4
Collected by Bipin Timalsina

You might also like