Presentation is loading. Please wait.

Presentation is loading. Please wait.

Natural Language Processing AI Revision Lee McCluskey, room 2/07

Similar presentations


Presentation on theme: "Natural Language Processing AI Revision Lee McCluskey, room 2/07"— Presentation transcript:

1 Natural Language Processing AI Revision Lee McCluskey, room 2/07 Email lee@hud.ac.uklee@hud.ac.uk http://scom.hud.ac.uk/scomtlm/cha2555/

2 Natural Language Processing Revision Overview -- make sure you can do can do the exercises -- revise / understand the slides -- focus on points given out in Easter email repeated below -- some questions will be given out in tutorials, otherwise you can ask me to go through anything then

3 Natural Language Processing Planning and Search Know about state space search, be able to describe the graphplan algorithm. Know the definitions of jargon to do with search / heuristics (e.g. complete, optimal, admissable, greedy..) Be able to describe some planning state space heuristics. Know how planning operator schema are represented in PDDL and how they work. Know about Knowledge engineering (eg how it differs from software engineering). Know why Prolog is useful for applications/ experimentation.

4 Natural Language Processing 2 player games Explain static vs dynamic Board Evaluation, and the idea behind game trees Know the Minimax algorithm / a- b pruning Know how to apply minimax and a – b pruning.

5 Natural Language Processing Machine Learning Know Fundamental Types / Classes of Learning Be able to compare machine learning in Neural Nets, RL and symbolic learning. Macro learning / weakest precondition – know how it works in detail. Be able to create simple macros from past solutions like in exercises using weakest precondition

6 Natural Language Processing Data Mining Be able to write down jargon definitions (Item, Apply Associative Classification, Do simple AC examples like in exercises.

7 Natural Language Processing Know the difficulties in NLP - illustrate these with example sentences, Show how it might be possible to overcome them. Explain typical steps in NLP. Be able to show how Prolog grammar rules work. Be able to draw a parse tree. Be able to explain the Turing test.


Download ppt "Natural Language Processing AI Revision Lee McCluskey, room 2/07"

Similar presentations


Ads by Google