New Semester
Started
Get
50% OFF
Study Help!
--h --m --s
Claim Now
Question Answers
Textbooks
Find textbooks, questions and answers
Oops, something went wrong!
Change your search query and then try again
S
Books
FREE
Study Help
Expert Questions
Accounting
General Management
Mathematics
Finance
Organizational Behaviour
Law
Physics
Operating System
Management Leadership
Sociology
Programming
Marketing
Database
Computer Network
Economics
Textbooks Solutions
Accounting
Managerial Accounting
Management Leadership
Cost Accounting
Statistics
Business Law
Corporate Finance
Finance
Economics
Auditing
Tutors
Online Tutors
Find a Tutor
Hire a Tutor
Become a Tutor
AI Tutor
AI Study Planner
NEW
Sell Books
Search
Search
Sign In
Register
study help
business
management and artificial intelligence
Paradigms Of Artificial Intelligence Programming Case Studies In Common Lisp 1st Edition Peter Norvig - Solutions
Given the differences between the architectures of modern computers and that of the human brain, what relevance does research into the physiological structure and function of biological systems have for the engineering of AI programspg23 Justify your answer.pg23
Although computing is a relatively new discipline, philosophers and mathematicians have been thinking about the issues involved in automating problem solving for thousands of years. What is your opinion of the relevance of these philosophical issues to the design of a device for intelligent problem
Describe your own criteria for computer software to be considered "intelligent."pg23
Criticize Turing's criteria for computer software being "intelligent."pg23
Discuss your thoughts on the mind-body problem and its importance for a theory of artificial intelligence.
Much traditional Western thought has dwelt on the mind-body relationship. Are the mind and body:a. b. distinct entities somehow interacting, or is mind an expression of "physical processes," or C. is body just an illusion of the rational mindpg23
Give several other examples of Aristotle's distinction between "matter" and "form." Can you show how your examples might fit into a theory of abstractionpg23
Answers that are neither exact nor optimal, but are in some sense "sufficient." This is a result of the essential reliance on heuristic problem-solving methods in situations where optimal or exact results are either too expensive or not possible.pg23
An attempt to deal with issues of semantic meaning as well as syntactic form.pg23
Reasoning about the significant qualitative features of a situation.pg23
A concern with problem solving using inexact, missing, or poorly defined information and the use of representational formalisms that enable the programmer to compensate for these problems. pg23
A focus on problems that do not respond to algorithmic solutions. This underlies the reliance on heuristic search as an AI problem-solving technique.pg23
The use of computers to do symbolic reasoning, pattern recognition, learning, or some other form of inference.pg23
Little learning from experience. Current expert systems are handcrafted; once the system is completed, its performance will not improve without further attention from its programmers. This leads to severe doubts about the intelligence of such systems.pg23
Difficulties in verification. Though the correctness of any large computer system is difficult to prove, expert systems are particularly difficult to verify. This is a serious problem, as expert systems technology is being applied to critical applications such as air traffic control, nuclear
Inability to provide deep explanations. Because expert systems lack deep knowledge of their problem domains, their explanations are generally restricted to a description of the steps they took in finding a solution. They cannot tell "why" a certain approach was taken.pg23
Lack of robustness and flexibility. If humans are presented with a problem instance that they cannot solve immediately, they can generally return to an examination of first principles and come up with some strategy for attacking the problem. Expert systems generally lack this ability.pg23
Difficulty in capturing "deep" knowledge of the problem domain. MYCIN, for example, lacks any real knowledge of human physiology. It does not know what blood does or the function of the spinal cord. Folklore has it that once, when selecting a drug for treatment of meningitis, MYCIN asked whether
3.6 that allows complex sentences, such as, sentence sentence pg24pg24 sentence.
Add grammar rules to Example
3.6. pg24
Add rules for (multiple) prepositional phrases to Example
3.6. pg24
Add rules defining adjectives and adverbs to the grammar of Example
3.5 for the case of an individual with four dependents, $18,000 in the bank, and a steady income of $25,000 per year. Based on a comparison of this problem and the example in the text, suggest a generally "best" strategy for solving the problem. pg24
Trace a data-driven execution of the financial advisor of Example
Give another example of an and/or graph search problem and develop part of the search space. pg24
3.4 in a data-driven fashion. pg24
Trace the good-dog problem of Example
Write a backtrack algorithm for and/or graphs. pg24
pg24
Choose and justify a choice of breadth- or depth-first search for examples of Exercise
Determine whether goal-driven or data-driven search would be preferable for solving each of the following problems. Justify your answer. pg24a. Diagnosing mechanical problems in an automobilepg24.b. C.d. e.f. You have met a person who claims to be your distant cousin, with a common ancestor named
Implement a backtrack algorithm in a programming language of your choice. pg24
27. Begin from state A. Keep track of the successive values of NSL, SL, CS, etc. pg24
"Hand run" the backtrack algorithm on the graph in Figure
Give an instance of the traveling salesperson problem for which the nearest-neighbor strategy fails to find an optimal path. Suggest another heuristic for this problempg24
Discuss the advan- tages of breadth-first and depth-first for searching this spacepg24
2). Let the nodes represent states of the world; e.g., the farmer and the goat are on the west bank and the wolf and cabbage on the eastpg24
1 and
3 (see Figures
Give the graph representation for the farmer, wolf, goat, and cabbage problem of Section
Is there such a path in the Knigsberg mappg24
A Hamiltonian path is a path that uses every node of the graph exactly once. What condi- tions are necessary for such a path to existpg24
How can an interpreter be designed to most effectively utilize a representation languagepg24
How can the interpreter most effectively reduce search complexitypg24
What is the complexity of the search process in terms of time usagepg24 Space usagepg24
When a solution is found, is it guaranteed to be optimalpg24
Will the problem solver always terminate, or can it become caught in an infinite looppg24
List and discuss two potentially negative effects on society of the development of artificial intelligence techniques.pg24
Add two more benefits for expert systems to those already listed in the text. Discuss these in terms of intellectual, social, or financial results.pg24
Pick one problem area that you feel would justify the energy required to design an expert system solution. Spell the problem out in some detail. Based on your own intuition, which aspects of this solution would be most difficult to automatepg24
Given the differences between the architectures of modern computers and that of the human brain, what relevance does research into the physiological structure and function of biological systems have for the engineering of AI programspg24 Justify your answer.pg24
Although computing is a relatively new discipline, philosophers and mathematicians have been thinking about the issues involved in automating problem solving for thousands of years. What is your opinion of the relevance of these philosophical issues to the design of a device for intelligent problem
Describe your own criteria for computer software to be considered "intelligent."pg24
Criticize Turing's criteria for computer software being "intelligent."pg24
Discuss your thoughts on the mind-body problem and its importance for a theory of artificial intelligence.
Much traditional Western thought has dwelt on the mind-body relationship. Are the mind and body:a. b. distinct entities somehow interacting, or is mind an expression of "physical processes," or C. is body just an illusion of the rational mindpg24
Give several other examples of Aristotle's distinction between "matter" and "form." Can you show how your examples might fit into a theory of abstractionpg24
Answers that are neither exact nor optimal, but are in some sense "sufficient." This is a result of the essential reliance on heuristic problem-solving methods in situations where optimal or exact results are either too expensive or not possible.pg24
An attempt to deal with issues of semantic meaning as well as syntactic form.pg24
Reasoning about the significant qualitative features of a situation.pg24
A concern with problem solving using inexact, missing, or poorly defined information and the use of representational formalisms that enable the programmer to compensate for these problems. pg24
A focus on problems that do not respond to algorithmic solutions. This underlies the reliance on heuristic search as an AI problem-solving technique.pg24
The use of computers to do symbolic reasoning, pattern recognition, learning, or some other form of inference.pg24
Little learning from experience. Current expert systems are handcrafted; once the system is completed, its performance will not improve without further attention from its programmers. This leads to severe doubts about the intelligence of such systems.pg24
Difficulties in verification. Though the correctness of any large computer system is difficult to prove, expert systems are particularly difficult to verify. This is a serious problem, as expert systems technology is being applied to critical applications such as air traffic control, nuclear
Inability to provide deep explanations. Because expert systems lack deep knowledge of their problem domains, their explanations are generally restricted to a description of the steps they took in finding a solution. They cannot tell "why" a certain approach was taken.pg24
Lack of robustness and flexibility. If humans are presented with a problem instance that they cannot solve immediately, they can generally return to an examination of first principles and come up with some strategy for attacking the problem. Expert systems generally lack this ability.pg24
Difficulty in capturing "deep" knowledge of the problem domain. MYCIN, for example, lacks any real knowledge of human physiology. It does not know what blood does or the function of the spinal cord. Folklore has it that once, when selecting a drug for treatment of meningitis, MYCIN asked whether
use the first method
Now, using the difference from the goal as the cost function, we can search using best-first search:> ( b e s t - f i r s t - s e a r c h 1
"Search lowest cost states f i r s t until goal i s reached."(tree-search ( l i s t s t a r t ) goal-p successors (sorter cost-fn)))
"Return the function that finds the difference from num."#'(lambda (x) (abs (- x num))))(defun sorter (cost-fn)"Return a combiner function t h a t sorts according to cost-fn."#'(lambda (new old)(sort (append new old) #'< :key cost-fn)))(defun best-first-search ( s t a r t goal-p successors cost-fn)
To implement best-first search we need to add one more piece of information: a cost function that gives an estimate of how far a given state is from the goal.For the binary tree example, we will use as a cost estimate the numeric difference from the goal. So if we are looking for 12, then 12 has
While breadth-first search is more methodical, neither strategy is able to take advantage of any knowledge about the state space. They both search blindly. In most real applications we will have some estimate of how far a state is from the solution. In such cases, we can implement a best-first
"Return a successor function t h a t generates a binary tree with n nodes."#'(lambda (x)(remove-if #'(lambda (child) (> child n))(binary-tree x) 1))> (depth-first-search 1 ( i s 12) (finite-binary-tree 15));; Search: (1);; Search: (2 3);; Search: (4 5 3);; Search: (8 9 5 3);; Search: (9 5 3);;
If the search tree is finite, then either breadth-first or depth-first will eventually find the goal. Both methods search the entire state space, but in a different order. We will now show a depth-first search of the 15-node binary tree diagrammed previously.It takes about the same amount of time
Breadth-first search ends up searching each node in numerical order, and so it will eventually find any goal. It is methodical, but therefore plodding. Depth-first search will be much faster-if it happens to find the goal at all. For example, if we were looking for 2048, depth-first search would
The only difference between depth-first and breadth-first search is the difference between append and prepend. Herewe see breadth-fi rst-searchinaction:> (breadth-first-search 1 ( i s 12) 'binary-tree);; Search: (1);; Search: (2 3);; Search: (3 4 5);; Search: (4 5 6 7);; Search: (5 6 7 8 9);;
The problem is that we are searching an infinite tree, and the depth-first search strategy just dives down the left-hand branch at every step. The only way to stop the doomed search is to type an interrupt character.An alternative strategy is breadth-first search, where the shortest path is
To make it easier to specify a goal, we define the function i s as a function that returns a predicate that tests for a particular value. Note that i s does not do the test itself.Rather, it returns a function that can be called to perform tests:(defun i s (value) #'(lambda (x) (eql x value)))Nowwe
"Search new states f i r s t until goal is reached."(tree-search ( l i s t start) goal-p successors #'append))Let's see how we can search through the binary tree defined previously. First, we define the successor function b i nary- tree. It returns a list of two states, the two numbers that are
The first strategy we will consider is called depth-first search. In depth-first search, the longest paths are considered first. In other words, we generate the successors of a state, and then work on the first successor first. We only return to one of the subsequent successors if we arrive at a
"Find a state t h a t satisfies goal-p. Start with states, and search according to successors and combiner."(dbg :search ""&;; Search: "a" states)(cond ((null states) fail((funcall goal -p ( f i r s t states)) ( f i r s t states))(t (tree-search(funcall combiner(funcall successors ( f i r s t
We will call our first searching tool tree-search, because it is designed to search state spaces that are in the form of trees. It takes four arguments: (1) a list of valid starting states, (2) a predicate to decide if we have reached a goal state, (3) a function to generate the successors of a
The successors, or states that can be reached from any other state.The strategy that determines the order in which we search.The first three features are part of the problem, while the fourth is part of the solution. In GPS, the starting state was given, along with a description of the goal states.
Search problems are called nondeterministic because there is no way to determine what is the best step to take next. A1 problems, by their very nature, tend to be nondeterministic. This can be a source of confusion for programmers who are used to deterministic problems. In this section we will try
6.4 A Set of Searching Tools The GPS program can be seen as a problem in search. In general, a search problem involves exploring from some starting state and investigating neighboring states until a solution is reached. As in GPS, state means a description of any situation or state of affairs. Each
"Find some rule with which to transform the input."(rule-based-trans1 ator input *el iza-rules*:action #'(lambda (bindings responses)(sublis (switch-viewpoint bindings)(random-elt responses)))))
( l e t ((result (funcall matcher (funcall rule-if rul el input)))(if (not (eq result f a i l ) )(funcall action result (funcall rule-then r u l e ) ) ) ) )rules))(defun use-eliza-rules (input)
"Find the f i r s t rule in rules t h a t matches input, and apply the action to t h a t rule."(some#'(lambda (rule)
a What to do when a rule matches. Once we have determined which rule to use, we have to determine what it means to use it. The default is just to substitute the bindings of the match into the then-part of the rule.The rule-based translator tool now looks like this:(defun rule- based-trans1
e How to see if a rule matches. By default, we will use pat -match, but it should be possible to use other matchers.
a What list of rules to use. In general, each application will have its own list of rules.
a What kind of rule to use. Every rule will be characterized by an if-part and a then-part, but the ways of getting at those two parts may vary.
It turns out that this will be a quite common thing to do: search through a list of rules for one that matches, and take action according to that rule. To turn the structure of use - el i za - rul es into a software tool, we will allow the user to specify each of the following:
As we have defined it, the pattern matcher matches one input against one pattern. In el i za, we need to match each input against a number of patterns, and then return a result based on the rule that contains the first pattern that matches. To refresh your memory, here is the function use-el i za -
6.3 A Rule-Based Translator Tool
p Exercise 6.2 [h] In the few prior examples, every time there was a binding of pattern variables that satisfied the input, that binding was found. Informally, show that pat -match will always find such a binding, or show a counterexample where it fails to find one.
Showing 1700 - 1800
of 4756
First
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Last
Step by Step Answers
Get 50% OFF
Claim Now
