1 / 23

Game AI versus AI (H. Mu ñ oz-Avila)

Game AI versus AI (H. Mu ñ oz-Avila). What is the difference between Game AI an AI? Can this difference be overcome? What is an attack Kung-Fu style? Discuss some alternatives to get the same effect Explain the difference between the traditional view to AI and the modern view.

stefan
Télécharger la présentation

Game AI versus AI (H. Mu ñ oz-Avila)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Game AI versus AI (H. Muñoz-Avila) • What is the difference between Game AI an AI? Can this difference be overcome? • What is an attack Kung-Fu style? Discuss some alternatives to get the same effect • Explain the difference between the traditional view to AI and the modern view. • Describe the Turing Test. Explain the mathematical objection against AI that follows from the Turing test

  2. East Coast Games • What is the difference between Angel Investment and Venture Capital? • Name two alternatives for incorporating companies • What does the speaker recommended: to make 10,000 or 1,000,000 stocks? Why?

  3. Finite Machines CH. 5.1 & 5.2 (Jarret Raim) • Give an example of how you would make use of the onEnter() and onExit() events in your FSM? • What is a stack based FSM and give an example of where it would be useful? • Why are Data Driven FSM's better than other implementations even though they require a lot of extra work? • Describe the Level of Detail optimization and explain why it works? • Describe online and offline debugging techniques. • Give an example where polymorphic FSM's would be useful and why.

  4. Finite Machines CH. 5.3 & 5.4 (Chad Hogg) • What is the difference between the definition of Finite State Machine from the book and the definition commonly provided in theory texts (e.g., finite automata) • What are the advantages / disadvantages of a Data-driven FSM implementation vs. a conventional FSM implementation? • What are the advantages / disadvantages of a Scripted FSM implementations vs. a conventional FSM implementation? • What is the difference between the data-driven and scripted models? • What are the three main types of data that must be specified in our implementation of an FSM?

  5. AI Planning (H. Muñoz-Avila) • What does the term “classical” planning refers to • What does the search space represents in planning • What is hierarchical planning? Give an example of a hierarchical planning process

  6. Autonomous AI Ch. 3.1, 3.2 (Stephen Hookway) • Name two problems with fully autonomous AI as it relates to gaming? • What is the method the designers used in No One Lives Forever 2 to give the appearance of coordinated movement? • Why is Transparency important in the MOUT situation but not in a gaming situation? • Why can’t you use a stack to place a dependency on two actions? • Name one use of the shared Blackboard

  7. Group movement, tactics and planning Ch. 3.3, 3.4 (Stephen Grabowski) and Additional Notes (H. Muñoz-Avila) • Name three advantages of using GOAP? • What is the Sussman anomaly and how does the planning system in FEAR avoids it? • Explain the tactics that a squad of NPCs should follow when firing upon there enemy. Draw a diagram illustrating these tactics • Explain in your own words why planning and FSMs are equivalent (don’t say “because they are Turing complete”) • Explain what the “smallest angle to clear view” method is pertaining to NPC movement?

  8. Hierarchical planning Ch. 3.5, 7.9 (Ke Xu) • What are the problems with the reactive approach of game AI? • What’s the basic idea of Hierarchical Planning? • List the major elements that Hierarchical Planning uses. • Describe two of these element briefly. • Tell the difference between strategy change and modification (re-planning).

  9. Pathfinding Ch. 2.1, 2.2 (Moll) •  For following search space representations, which would you usein which environment (open world or human architecture) and why?a.        Grid-Basedb.        Waypoint Graph Basedc.        Navigation Meshes • Discuss complexity of the search procedures for the following search spacerepresentationsa.        Grid-Basedb.        Waypoint Graph Basedc.        How do Navigation Meshes, Circle Waypoint Graphs, etc improveupon complexity of general waypoint graphs? • Which representations work best with the local pathfinding system and obstacle avoidance?  Why? • What is a navigation set? •  For the precomputed pathfinding solution highlighted in Section 2.2 what part (which set) does the cost arise from and why? • What type of world or situation would you use something like potential fields to plan movement?

  10. Pathfinding Ch. 2.3, 2.4 (Hoang) • What are the advantages of look-up tables over searching algorithms such as A*? Why would someone consider using A* instead of look-up tables? • Describe how does the indexed look-up tables are implemented. What is it'smemory usage if there are N waypoints (describes in term of N, assume each waypoint will take up 2 bytes)? • Why is area-based look-up tables usually preferred over regular matrix look-uptables and indexed look-up tables? • Describe reactive and deliberative planning.  Why is it a good ideato use the hybrid planning? • Reactive planning has difficulties with integration of multiple behaviors. It hasbeen suggested that by prioritizing the behaviors, it can fix this problem. Describe how prioritizing would fix the problem.  In order words, how does it work?

  11. Pathfinding 2.5 & 2.6 (Scott Lindner) 1.  How can you utilize your current navigation system to aid actionssuch as grenade throwing, or when/where to climb a ladder?2.  Using embedded information, explain how to determine whether or notan obstacle is big enough to hide behind.3.  Why is it a bad idea when an NPC is chasing after you for it to alwaystravel to the location where you currently are located? 4.  What are the 3 scenarios you must consider when trying to have an NPChunt down a player in a convincing manner?5.  Give an example of using edge-cost biasing to influence pathfinding.

  12. FPS, RTS and Strategy. Ch. 7.1, 7.2 (Christopher Mansley) • What motivations are there for qualitative type reasoning verses quantitative? • What are the two types of spatial reasoning discussed? • Describe the openness layer and how it is used. • What is one type of analysis (or extracted information) performed by Masters of Orion 3? • Why is a qualitative reasoning more useful than a simple path finding? • Name one technique used in Thief to allow the NPCs to reason about their environments.

  13. FPS, RTS and Strategy. Ch. 7.4, 7.5 (Stephen-Lee Urban) • Why are script files and a separate .dll for the map engine a good design strategy? • List and explain 3 reasons why player placement is so important. • What are the two steps in map generation requiring an elevation filter, and how might this filter operate? • What challenges would the creation of transports for flying units pose (i.e., Relax restriction that only ground units are transported; e.g. aircraft carrier)? • Why is a line of fire, fixed-obstacle check used as the “last-chance” verification that a very close unit can or cannot board a naval transport?

  14. FPS, RTS and Strategy. Ch. 7.6, 7.8 (Patrick Schmid) • Why is a wall-building algorithm based on the idea of expanding the wall not practical? • Describe an efficient wall-building algorithm • How would the presented wall-building algorithm incorporate gates ordoors? • What are the benefits of a multi-tiered AI system? • Describe a combat scenario that a 4-tiered AI system would make possible. Highlight what each tier would know about the world and what kinds of orders it would execute. Assume that the 4 tiers are: computer player AI, platoon, squad, soldier

  15. Scripting Ch. 9.1, 9.2, 9.3 (Ian Warfield) • List some advantages and disadvantages of scripting. • Why might designing your game to be “scriptable” be a smart business move? • When might it be a good idea to design a game with Level 1 scripting (data specified in files)? (Hint: In what circumstances does this kind of game scale well?) Give an example of such a game. • Why might a game developer prefer to use scripted cutscenes instead of pre-rendered cutscenes? • Compare and contrast the declarative and imperative scripting models. Why might a mission/map/level designer prefer one over the other?

  16. Racing and Sports. Ch. 8.1, 8.2, 8.3 (Upmanyu Misra) 1) Brooks' Subsumption architecture is 'purely reactive'. This means that layers involved do not keep track of history. Downforce AI is based on Brooks' Subsumption, but is not purely reactive. Why? 2) Briefly describe i) the four layers used for Downforce AI, ii) the behaviors (sub-tasks) of each layer 3) How does the Downforce AI address the issue of optimal positioning of a vehicle on the racing track? 4) A vehicle needs to make a pit stop for a change of tyres due to wearing off which is evident from excessive skidding of tyres. The pit stop lane will start immediately after the apex of an approaching corner. i) Which layers of the Downforce AI design will play a role in the whole process. ii) What will be the hierarchy between those layers if order of execution is the criteria for awarding control to a layer.

  17. Racing and Sports. Ch. 8.4, and 8.5 (Ushhan D. Gundevia) • Discuss one of the approaches for DR that the author talks about? • Explain the Cubic Spline method for DR described by Nick Caldwell. Contrast with any three methods described by the author? • Explain the Targeting method used to overcome lag in online games. • What are the unique challenges faced while designing AI for Sports Games? • For every Agent Plan in the architecture, we have the following functions • EvaluateInitiation() • EvaluateContinuation() • Initiate() • Update() What are they used for? Two functions return float values. Which ones and why? 6. Give a brief description of the Shot Plan.

  18. Machine Learning: A Quick Look (H. Muñoz-Avila) • With the same set-up as the example we saw in class: suppose that the only positive instance is: obj(large, white, ball) and the only negative instance is: obj(small,red,brick). Give an example of a valid generalization and of an invalid generalization for these instances • What is the purpose of using Explanation-Based Learning in Planning? • Describe the genetic algorithm. Give an example of possible application of the genetic algorithm. What would be the fitness function in your example

  19. Learning + User Modeling. Ch. 10.1 (Christopher Janneck) 1. What is the simplest form of player model?  What are some benefits of this type?2. How can you make a more advanced player model?  What does your design afford you that the simple model does not?3. Give 2 additional ways that a player model can be used with a game, outside that of improving or adjusting the AI.4. Provide an example when modifying the AI online is more (or equally) effective than making modifications offline?  And vice versa?5. Do you think that the Robocup project will achieve its goal (develop a team of autonomous robots that will beat the World’s Championship Soccer (Football) team by 2050)?  How do you perceive player modeling to fit into your prediction?

  20. TIELT (Testbed for Integrating and Evaluating Learning Techniques) (David W. Aha) • What are the limitations of current research on Machine Learning? • What is it that the Tielt project is trying to achive? • What are the potential benefits of Tielt for a Machine Learning researcher? What are the potential benefits for a game developer? • Discuss the Tielt architecture (you don’t need to list all components, nor to remember their exact names) • What do you think are the main obstacles of constructing a system like Tieet?

  21. Learning + decision trees. Ch. 10.2 (Christopher Creswell) • What is the idea of the ID3 algorithm • Name two of the three conditions that stop the recursion in learning a decision tree from a set of examples. • What method does the ID3 algorithm use to pick the best attribute to classify the examples on (you don't have to go into detail, a brief description will suffice)? • Name one weakness of the ID3 algorithm • What extension to the ID3 algorithm did we discuss to make it work well with continuous values?

  22. Learning + machine learning experiment. Ch. 10.4 (Marc Ponsen) • What is online learning? Name a (possible or real) application for online learning in computer games • Discuss two of the four requirements for online use of a learning technique in computer games? • Why is the Dynamic Scripting technique frequently delivers effective behavior? • What is offline learning? Name a (possible or real) application for online learning in computer games • Which machine learning technique was discussed to be effective in searching for new tactics in RTS games?

  23. learning and recognition. Ch. 10.3, 10.5 (Kit Ming Chan) • Give some examples of possible use of pattern recognition in gaming • What’s the difference between pattern recognition as an optimization and an adaptation problem? • What is the idea behind the formulation of the levels of decision making? • Name two of the four questions to determine if machine learning is suitable for a certain game? Illustrate your answers with Black and White. • When does instance-based learning learn the wrong knowledge?

More Related