"AI" course exercises _ AI

Source: Internet
Author: User
Tags reflection
Introduction to the first chapter of the "AI" course exercises

1-1. What is AI. Try to explain from two aspects of discipline and ability.

1-2. In the development of artificial intelligence, what thoughts and thoughts played an important role.

1-3. Why can the machine (computer) imitate human intelligence.

1-4. What are the schools of artificial intelligence now? What are their cognitive perspectives.

1-5. What levels do you think should be studied on cognitive behavior?

1-6. What are the main areas of research and application of artificial intelligence. Among them, which is the new research hotspot.

Chapter II Knowledge Representation method

2-1 the key points of state space method, problem attribution constitution, predicate logic method and semantic network method. What are the essential connections and similarities and differences between them?

2-2 3 Missionaries and 3 savages came to the river, intending to ferry from the right bank to the left bank by a ship. The load capacity of the ship is two people. At any time, if the number of savages exceeds the number of missionaries, then the savage will eat the missionaries. How can they use the ship to safely cross all the rivers?

2-3 using FIG. 2.3, a shortest travel distance is planned with the state space method: This journey begins with City A, visits other cities not more than once, and returns a. Select a state representation that represents the node and arc of the state space being obtained, marking the appropriate price and indicating the optimal path from the starting node to the target node in the graph.

2-4 try to show how to use a tree or a tree to express the calculation of the electrical network impedance shown in Fig. 2.28. The fact that separate R, L, or C can be computed separately with R, Jωl, or 1/jωc is used as a primitive problem. The successor operator should be based on the rules of composite parallel and series impedance.

Figure 2.28

2-5 try the structure of four-yuan series to represent the four-disc van Gogh problem, and draw a diagram to solve the problem.

2-6 transform the following sentences into clause forms:

(1) (x) {P (x) →p (x)}

(2) XY (On (x,y) →above (x,y))

(3) XYZ (Above (x,y) ∧above (y,z) →above (x,z))

(4) ~{(x) {p (x) →{(y) [P (Y) →p (f (x,y))]∧ (y) [Q (X,y) →p (y)]}}

2-7 use predicate calculus formula to express the following English sentences (multiple use rather than province with different predicates and items.) For example, do not use a single predicate letter to represent each sentence. )

A computer system is intelligent if it can perform a task which,if performed by a human, requires intelligence.

2-8 The following statements represent an idiom semantic network description:

(1) All Mans are mortal.

(2) Every cloud has a silver lining.

(3) All branch managers of DEC participate in a profit-sharing plan.

2-9 as a movie audience, please write a script to go to the cinema to see a movie.

2-10 try to construct a frame system that describes your dormitory or office. Chapter Three Search Inference technology

3-1 What is the graph search process? Where does the rearrangement of the open table mean, and what is the principle of rearrangement?

3-2 tries to compare the efficiency of various search methods.

3-3 What are the steps in the form of clauses? Please combine examples to illustrate this.

3-4 How to find the answer to the problem by digestion and inversion?

3-5 what is the appropriate formula? What is the equivalence relation of the appropriate formula?

3-6 Use width first search to find the way out of the maze shown in Fig. 3.33.

Fig. 3.33 A case of maze

3-7 using bounded Depth-first search method to solve the eight digital problems shown in Figure 3.34.

















So Sg

Figure 3-34 Eight digital puzzle

3-8 use the latest methods to express missionaries and savages, and write a computer program to secure the solution of all 6 people.

Tip: When applying state space representation and search methods, you can use (NM,NC) to represent the status description, where Nm and Nc are the number of missionaries and savages respectively. The initial state is (3,3), while the possible intermediate states are (0,1), (0,2), (0,3), (1,1), (2,1), (2,2), (3,0), (3,1) and (3,2), etc.

3-9 try to compare the search efficiency of breadth first search, bounded depth first search and ordered search, and illustrate with instance data.

3-10 Robots drive trucks and carry parcels (numbered, #3, respectively) to the 3 residences of Lin (Lin), Wu and Hu. Some simple operators are specified, such as drive (X,y), which represents driving directions, and Unload (z)  for unloading the parcel, and there are certain prerequisites and results for each operator. The test shows how the state space problem solving system can derive an operator sequence by using predicate calculus, which can generate a satisfying at (#1, LIN) ∧at (#2, WU) ∧at (#3, HU) and target state.

3-11 rule deduction system and production system what kinds of reasoning methods? What are their characteristics?

3-12 Why is it necessary to adopt system organization technology? What kinds of systems organize technology?

3-13 What is the significance of studying uncertain reasoning? What are the uncertainties?

3-14 What are the limitations of monotone inference? What is the default inference? How does a non-monotone inference system Verify the validity of a node?

3-15 Under what circumstances is there a need for uncertain or non monotone reasoning?

3-16 The following statements are geometric theorems that represent the production rules for rule-based geometric proof systems:

(1) The corresponding angles of the two congruent triangles are equal.

(2) The corresponding edges of two congruent triangles are equal.

(3) The triangles equal to each corresponding edge are congruent triangles.

(4) The two corners of the isosceles are equal.

Fourth chapter Computational Intelligence (1): Neuro-Computational fuzzy computation

4-1 What is the meaning of computational intelligence. What branches of research it involves.

4-2 The relationship between computational intelligence (CI), artificial intelligence (AI) and biological Intelligence (BI) is discussed.

4-3 why the artificial neural network has the attractive development foreground and the potential widespread application domain.

4-4 the structure and main learning algorithms of biological neuron and artificial neural network are described.

4-5 consider a neural network with a ladder-type threshold function, assuming

(1) Multiply all weights and thresholds with a constant;

(2) A constant is added to the ownership value and the threshold value.

Try to show whether network performance will change.

4-6 is used as a neural network to compute an XOR function that contains 2 inputs. Specifies the type of neural network unit used.

4-7 assumes that there is a neural network with a linear excitation function, i.e., for each neuron, its output equals constant c multiplied by the input weighted sum.

(1) There is an implied layer of the network. For a given power W, the output value of the output-layer unit is written out, which is based on the weight w and input layer I, and there is no explicit description of the output of the hidden layer. The test proves that a network without implied units can compute the same function.

(2) Repeat the above calculation for a network with any hidden layer number. The conclusion of linear excitation function is given.

4-8 the data structure of a layered feedforward neural network is implemented to provide the necessary information for forward evaluation and reverse propagation. Using this data structure, write out a neural network output as an example and compute the appropriate output value for the network.

4-9 what is fuzziness. What is the opposite meaning of it. Try each of the two examples to illustrate.

4-10 What is a fuzzy set and membership function or membership degree.

4-11 What are the operations of fuzzy sets and what laws are met.

4-12 what is fuzzy inference. What kinds of fuzzy inference methods are available.

4-13 What are the fuzzy implication relationships.

4-14 What is called fuzzy judgment. What are some common methods of fuzzy judgment.

4-15 the quality of a product for a random assessment. 5 products are randomly selected x1,x2,x3,x4,x5 for inspection, and their quality is as follows:


This determines a fuzzy set Q that represents the degree of subordination of the fuzzy concept of "quality level" for the group of products.

Try to write the fuzzy set.

4-16 has the following two fuzzy relationships

The compound relation r10r2 between R1 and R2 is tried out.

Fifth chapter Computational Intelligence (2): Evolutionary computation of artificial life

5-1 what is evolutionary computing. What it includes. What their starting point is.

5-2 the basic principle of genetic algorithm is discussed and the solving steps of genetic algorithm are explained.

5-3 How to use genetic algorithm to solve the problem, try to illustrate the solution process.

5-4 maximum value by genetic algorithm

5-5 How the evolutionary strategy is described.

5-6 describes the mechanism and basic process of evolutionary programming, and illustrates the representation of evolutionary programming with four-state machines as an example.

5-7 the relationship between genetic algorithm, evolutionary strategy and evolutionary programming. What's the difference.

5-8 whether artificial life has been studied since 1987. Why.

5-9 what is artificial life. Please define the artificial life in your own language as you understand it.

5-10 Artificial life to imitate the characteristics and phenomena of natural life. What are the common characteristics of natural life?

5-11 Why to study artificial life.

5-12 What research content is included in artificial life. How to study the method. Sixth chapter expert System

6-1 What is an expert system? What are its features and advantages?

6-2 What are the components of the expert system and what are the roles of the components?

6-3 what are the key steps in building an expert system?

6-4 What is the difference between the expert System program and the general Problem solving software program? How is the task of developing expert systems different from the development of other software?

6-5 rules based expert system is how to work. Its structure.

6-6 framework-based expert systems are related to goal-oriented programming. What are the characteristics of the structure? What is its design task.

6-7 why to propose a model-based expert system. The general structure of neural network expert system is discussed.

6-8. What are the characteristics of the new expert system? What is distributed expert system and cooperative expert system.

6-9 What techniques should be considered when designing an expert system.

6-10 What is the tool for building an expert system? Do you know what expert system tools, what are the characteristics of each?

6-11 What is the problem with expert systems? How do you think the expert system should be developed?

6-12 using rule-based reasoning system to prove the correctness of the following reasoning:

Dogs are known to bark and bite.

When any animal barks, it always noisy.

The Hound is a dog.

Conclusion the hounds are noisy.

Seventh Chapter Machine Learning

7-1 What is learning and machine learning? Why Study machine learning?

7-2 try to describe the basic structure of machine learning system, and explain the function of each part.

7-3 try to explain the mode of mechanical learning. What are the important problems in mechanical learning that need to be studied?

7-4 try to explain the pattern and learning method of inductive learning.

7-5 what is analogical learning. Its reasoning and learning process.

7-6 try to explain the basic principle, learning form and function of learning.

7-7 tries to illustrate the main idea of symbolic system and connection mechanism in machine learning.

7-8 in C language to write a set of computer programs for the implementation of BP learning algorithm.

7-9 try to solve the problem of salesman travel by using neural network model.

7-10 consider a neural network with a ladder-type threshold function, assuming

(1) Multiply all weights and thresholds with a constant;

(2) A constant is added to the ownership value and the threshold value.

Try to explain whether the network performance will change?

7-11 is the weight gain the ability to slow down BP learning?

7-12 what is knowledge discovery. What is the relationship between knowledge discovery and data mining?

7-13 try to explain the process of knowledge discovery.

7-14 What are some of the more common methods of knowledge discovery. Try to introduce it slightly.

7-15 What are the areas of application of knowledge discovery. The prospect of the development and application of knowledge discovery is prospected.

Eighth Chapter robot Planning

8-1 What are some of the most important robot high-level planning systems? What are their characteristics? What kind of planning method do you think has a great future?

8-2 let right (x), leave (x), Up (x) and down (x) represent the units on the left, right, top, and bottom of Unit x in the eight-digit puzzle (if such a unit exists). Try to write stips plan to simulate moving up B (space), moving down B, moving B to the left, and moving B to the right.

8-3 consider designing a cleaning kitchen planning problem.

(1) write a set of strips operators that may be used. When you describe these operators, consider the following:

• Cleaning the floor with a stove or fridge.

• To clean the oven, you must apply the oven cleaner, and then remove the cleaner.

• Clean the floor before cleaning it.

• Before cleaning the floor, you must first take out the trash cans.

• Cleaning the refrigerator to create garbage and dirt, and the work table dirty.

• Cleaning the table or floor to make washing gossip dirty.

(2) Write a description of the possible initial state of a swept kitchen and write a descriptive (but potentially difficult to obtain) target description.

(3) Explain how to use strips planning technology to solve this problem. (Hint: You may want to revise the definition of adding a condition so that when a condition is added to the database, it can automatically delete the negation if it appears to be negative).

The volume of F1 and F2 of the 8-4 curved neck bottles were C1 and C2 respectively. The formula Cont (X,Y) represents a liquid in which the bottle X contains a unit of Y capacity. Try to write a strips plan to simulate the following actions:

(1) pour all the liquid in the F1 into the F2.

(2) Fill the F2 with the F1 part of the liquid.

8-5 Robot Rover is outside the room, want to enter the room, but can not open the door to let themselves in, and can only shout, let the call prompted the door. Another robot Max is in the room, he can open the door and like peace. Max can usually open the door to stop the rover from shouting. Suppose Max and Rover each have a strips planning generation system and a planning execution system. Try to explain the strips rules and actions of Max and Rover, and describe the planning sequence and execution steps that lead to a balanced state.

8-6 using any of the planning generation systems discussed in this chapter, solve the problem of stacked wood in the manipulator shown in Fig. 8.22.

8-7 Consider the search path problem shown in Figure 8.23.

(1) Establish a structural space for the object and the obstacle (shaded part). Where the initial position of the object has two kinds of conditions, one as shown, the other is to rotate the object 90 °.

(2) Applying the structure space, describes a process (program) that seeks the above collision-free path to confine the problem to a two-dimensional problem without rotation.

(a) initial layout (b) Target layout

Fig. 8.22 planning problem of manipulator stacking wood

8-8 indicate the path to the problem of figure 8.23 in your process structure space, and describe how you can generalize the conclusions you obtained in the above question to include rotation.

Figure 8.23 A finding path problem

8-9 Figure 8.24 shows the world model of robot work. The robot is required to move 3 boxes BOX1, BOX2 and BOX3 to the target location as shown in Fig. E23 (b), and to set up the plan by using the expert system method and robot the planning sequence.

(a) The initial world model M0 (b) The target World model G0

Fig. 8.24 robot planning for moving boxes in one place

8-10 Figure 8.25 shows the world model of robot work. The robot is required to move the box from the room R2 initial position to the room R1 target location. The expert system of robot planning is established and the planning result is given.

Figure 8.25 Robot plan for moving from one room to another

Nineth Chapter Agent (Aigen body)

9-1 What is the characteristic of distributed artificial intelligence system? Compare the characteristics of the multi-ai real-body system.

9-2 What is the Aigen body. What is your opinion on the translation of the agent?

9-3 the characteristics of the structure of the Aigen body. How to classify the structure. What are the characteristics of each of these structures?

9-4 Aigen Why the body needs to communicate with each other.

9-5 the steps, types and modes of communication of Aigen body are discussed.

9-6 the main communication languages of the Aigen body. They all have their own characteristics.

What are some of the basic models of the 9-7-ai real-body system? What kinds of architecture are there?

9-8 try to illustrate the collaboration method, negotiation technique and coordination mode of the multi-ai real body.

9-9 Why do I need to learn and plan.

9-10 What do you think the research direction of the multi-AI system should be? And what is the prospect of its application.

9-11 Choose a field you are familiar with and write a page program to describe the role of Aigen and the environment. Indicates whether the environment is accessible, deterministic, episodic, static, and continuous. For this field, what kind of Aigen body structure is used is good.

9-12 design and implement several Aigen body with internal state, and measure its performance. For a given environment, how these Aigen approach the ideal body.

9-13 change the shape of the room and the location of the furnishings, add new furniture. Try to measure the Aigen in the new environment and discuss how to improve its performance in order to deal with more complex landforms.

9-14 some Aigen, once they learn a new sentence, immediately make inferences, while others can only make inferences when they are asked to do so. What is the difference between the two reasoning methods in the knowledge layer, the logical layer and the execution layer?

9-15 using Boolean circuits to design a logical Aigen for the unknown Provence World. The circuit is a collection of logic gates connecting input (sensing valves) and output (action valves).

(1) Try to explain why a trigger is needed.

(2) Estimate how many logic gates and triggers are needed.

Tenth Chapter Machine Vision

10-1 You can use the generalized cone language to describe a wedge-shaped object as a triangle with a certain size that moves along a straight axis. Please give a different description.

10-2 (1) In addition to surface normal (p,q,-1), there are two other vectors of interest: a vector pointing to a light source, which corresponds to some particular p and Q values, which are recorded as PS and (s for hypothetical daylight), a vector pointing to the sun (ps,qs,-1), and another pointing to the observer, the vector (0,0,- 1).

Using surface normal, solar vector and observation vector, some formulae related to the ejection angle, angle of incidence and phase angle can be obtained by P and Q. The following formula is proved to be true:

(2) pair and derive similar formulas.

10-3 known lambert the surface brightness equals. If the light source is right behind the observer, that is, the brightness corresponding to p and Q can be:

When it is a constant, the brightness e is a constant value. Because it is the equation of a circle on the plane PQ, we can conclude that the luminance line on the PQ reflection chart is a circular line when the light source is behind the observer. Try to prove that the shadow line is straight.

10-4 fixed a basketball or other spherical object, and tested it under a single small light source in the room. The light source is behind the observer.

(1) How does the brightness of the spherical light change?

(2) Why does the full moon look flat?

10-5 consider having a long cube flat above the Lambert wall, as shown in figure (a). The intensity of light along the AB line is roughly the same as that of Fig (b), and when the corner of the cube is a sleek transition, its light intensity is shown in figure (c).

10-5 Figure Lambert Cube and its light intensity distribution map

(1) in the PQ space, the exact position of the surface normals of each visible side of the cube is pointed out.

(2) in the PQ space, pointing to the light source direction, pointed out the desirable position.

(3) If the junction is mutation, try to draw the light intensity distribution along the CD line.

(4) Assuming that the junction is slick, try to draw the light intensity distribution along the CD line.

10-6 The following array represents the PQ projection of the dot matrix on an aerial photo image and the chain code of the observed brightness er:

-1-1 0.23 +1-1 0.23 +1-1 0.17

-1-10.23 +1-1 0.17 0 0 0.3

0 0 0.3 0 0 0.3 0 0 0.3

Suppose the observed brightness, wherein, corresponds to the luminance line of the light source directly on the Lambert reflection diagram behind the Observer. Try to classify each point image as stone, tree and headstone, assuming that their reflectance coefficients are 0.7, 0.5 and 0.3 respectively.

10-7 the reflection coefficient of a cover plate surface is not known. At different times from 3 separate light sources to the surface illumination. For the surface, the 3 light sources on this surface reflect the figure shown in the figure. The brightness observed when the light is irradiated separately is:

Question 10-7 Fig. 3 Reflection graphs

(1) A line in the PQ space that represents the trajectory when equal to 2, 3, and 4. Likewise, draw the lines that are equal to 0.5, 1 and 2 o'clock.

(2) Seek

10-8 to quantify the objects shown in the picture as 32x32 (graph paper-owned)

Question 10-8 figure need to digitize the object

(1) Create two images, each containing 3 objects. The objects in the two frames are required to have different dimensions, positions and directions.

(2) Calculating the moment of each of the 6 objects in two frames.

(3) Calculate the moment heart of each object.

(4) Calculate the center moment, nominal center moment and invariant moment of each object, and discuss the results.

(5) Calculate the shape coefficients of 6 objects, and discuss the results obtained.

10-9 Why does the Consight system use 2 light sources instead of 1 light sources?

10-10 in connectivity analysis, the segmentation of adjacent 2 rows is defined as the following 3 species:

Condition 1 does not overlap

Zero in the middle or more columns



Condition 2 does not overlap

Zero in the middle or more columns



Situation 3 Overlapping

is different from the situation 1, but also different from the situation 2.

The zone-closing rule is that a higher number replaces the lower number (except for the background "0").

(1) Scan the following 8x8 binary images from left to right line by row (b is background in figure). The number of elements on the image matrix after the connected domain is combined is pointed out as a solution to the connectivity analysis:

1 2 3 4 5 6 7 8

B-b b b b b B. B. b b

1 B 0 0 0 1 1 1 0 0 B

2 B 1 1 0 0 1 1 1 0 b

3 B 0 1 0 1 1 0 1 1 b

4 B 0 1 1 1 1 0 0 1 B

5 B 0 1 1 1 0 0 0 1 b

6 B 0 1 0 1 1 0 1 1 b

7 B 0 1 0 0 1 1 1 0 b

8 B 0 0 0 0 0 1 0 0 B

B-b b b b b B. b b

(2) Determine the scanning width of image coding in the subject (1).

11th Chapter Natural Language Understanding

11-1 what is language and language comprehension? What are the levels of natural language understanding and what are the functions of each level?

11-2 What is the relationship between natural language understanding and automatic language generation? What do you have in common when studying the two?

11-3 the ambiguity of language can be found at all levels: Word formation, speech, syntax and semantics. Try each example to illustrate.

11-4 write out the transfer networks that correspond to the following context-independent grammars:


Np→adjective Noun

Np→determiner Noun PP

Np→determiner Noun

Vp→verb adverb NP


Vp→verb adverb

Vp→verb PP

Pp→proposition NP

11-5 consider the following sentences

The old man′s glasses were filled with sherry.

Choose the word glasses what information is needed? What information means inappropriate meaning?

11-6 consider the following sentences:

Put the "Red block" on the "Blue Block" on the table.

(1) Write all valid syntactic analyses of the sentence in accordance with the syntactic rules.

(2) How to choose the appropriate meaning of the order by means of semantic information and environmental knowledge?

11-7 The syntax analysis tree is given for each of the following statements:

(1) David wanted to the movie with Linda.

(2) David wanted to the movie with Georgy William.

(3) He heard the story listening to the radio.

(4) He heard the boys listening to the radio.

11-8 Consider the problem of English conversation between a user and an interactive operating system.

(1) write out the semantic grammar to determine the language used in the conversation. These languages should ensure basic operations, such as describing events, copying and deleting files, compiling programs, and retrieving file directories.

(2) Use your semantic grammar to analyze the following sentences:

Copy from the new test MSS into the old test MSS.

Copy to old Test MSS out of new test MSS.

(3) The above two statements are analyzed with standard English grammar, and the grammatical fragments are listed.

(4) What are the differences between the above (2) and (3) grammars? What does this difference have to do with the difference between syntactic and semantic grammar?

11-9 A university has developed a database of student status management. Try to write a matching sample that is suitable for querying the contents of the database.

11-10 try to design a natural language question answering system in a specific application area.

12th Chapter Intelligent Control

12-1 Why is intelligent control an important new field of artificial intelligence?

12-2 What are the structural theories of intelligent control? What are their central ideas and content? Compared with traditional control, what are the characteristics of intelligent control?

What are the key points of hierarchical intelligent control for 12-3 Saridis? What are the functions at all levels? How to use entropy to measure the function of all levels?

12-4 What characteristics should be considered when designing an expert controller? What is the general structure model of the expert control system?

12-5 What is a learning control system? What are the research topics? What are the design principles of learning control systems?

The structure principle and control rule of fuzzy controller are explained in 12-6 test. What are some design methods of fuzzy controller?

The 12-7 set theory fields X and y are all finite fuzzy sets, respectively

The fuzzy matrix R represents a fuzzy relationship from X to Y. What is the meaning of the element rij of the fuzzy matrix R?

12-8 when the input precision is converted to the fuzzy quantity (fuzzy) in the working process of the fuzzy controller, the output is

The fuzzy quantitative change is the exact quantity (not fuzzy). What is the effect of these transformations?

12-9 What are the characteristics of an artificial neural network that makes it suitable for control? What kinds of neural controllers, their structure

And the principle of action?

12-10 What are the areas of application of intelligent control? Try to cite an example of an intelligent control application that you know better or be familiar with, and explain its working principle and control performance. 13th Chapter Outlook

13-1 How do you evaluate the development and controversy of artificial intelligence? What is the relationship between controversy and development?

13-2 What are the arguments of different schools of artificial intelligence on theories, methods and techniques?

13-2 the influence of the development of artificial intelligence on human beings. Try to understand the situation of their own understanding, from the economic, social and cultural aspects of how to explain.

13-4 a review of the future development of artificial intelligence.

13-5 What advice do you have for "artificial intelligence" or "intelligent system" courses and their teaching?


Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.