Learning
Download as PPT, PDF6 likes9,522 views
The document discusses different types of learning strategies and methods. It defines learning as the acquisition of knowledge through study. It describes several learning methods including rote learning, direct instruction, analogy, induction, and deduction. Rote learning involves simple memorization while direct instruction involves being told information. Induction involves forming general concepts from examples and deduction uses existing knowledge to derive new facts. The document provides examples to illustrate each learning method.
![EXAMPLE : Training Example Owner(Object,Aakash) & Color(Object,Brown)& Is(Object,Light) & Has(Object,Coffee) & HasPart(Object,Concavity) & ...... PRUNING Domain Knowledge Is(X,Light)& HasPart(X,Y) & Is(Y,Handle) -> Liftable (X) HasPart(X,Y) & Is a(Y,Bottom) & Is(Y,Flat) -> Stable (X) HasPart(X,Y) & Is a(Y,Concavity) & Is(Y,UpwordPointing)->OpenVessel (X) Goal Concept [ CUP ] X is a Cup iff X is Stable , Liftable, OpenVessel Operationality Criterion Liftable --- Light Stable ---- Flat Open Vessel --- Concavity [ CUP ]](https://image.slidesharecdn.com/learning-1319518515471-phpapp02-111025000949-phpapp02/85/Learning-15-320.jpg)
Learning
- 1. Learning Dictionary Def n : Knowledge Acquired by Study.
- 2. According to Simons: # Changes in the System in that are adaptive in the sense that they enable the system to do the same task or task drawn by the same population more efficiently the next time. # Process by which one entity acquires knowledge.
- 3. Learning Strategies: Learning by Advices. Learning by Examples. Learning by Problem Solving.
- 4. Learning Methods: MEMORIZATION( Rote Learning) DIRECT INSTRUCTION( By Being told ) ANALOGY INDUCTION DEDUCTION
- 5. Rote Learning( Memorization) Simple Requires least amount of interference Simply coping the knowledge in the same form that it will be used directly into knowledge base. Ex : used for memorizing multiplication table In case of data caching , we store computed values we do not have to recompute them later, thus saving a significant time .Caching is used in AI programs to produce some surprising performance improvements .Such caching is known as Rote Learning .
- 6. DIRECT INSTRUCTION (By being told) Slightly complex from previous. Requires more inference. Knowledge must be transformed into an operational form being integrated into knowledge base. EX : As teacher presents a number of facts directly to us in well organized manner.
- 7. ANALOGY Requires more inferencing Process of learning new concept or solutions through the use of similar known concepts or solutions. EX : when we learn to drive a truck using our knowledge of car driving.
- 8. INDUCTION: Similar to analogy learning Requires more inferring than first two methods. Used when we formulate a general concept after seeing a number of instances of examples of the concept. EX : We learn the concept of color or sweet taste after experiencing the sensations associated with the several examples of colored objects or sweet foods.
- 9. New facts are derived from the known facts. Requires more inference EX: we could learn deductively that Bob & Sam are cousin’s ,if we have the knowledge of Bob and Sam’s parents and rules for the cousin relationship. DEDUCTION :
- 10. ROTE LEARNING: When a computer stores a piece of data it is performing a simple form of learning. Uses the concept of Data Caching.
- 11. Game Tree : A D B E L O M P C H F I N W G J K Q S R T A U V Stored score A=10 A=10
- 12. INDUCTION LEARNING: Definition: Process of acquiring generalized knowledge from examples or instance of some classes. Terms: CLASS OBJECT CONCEPT HYPOTHESIS TARGET CONCEPT POSITIVE INSTANCE NEGATIVE INSTANCE
- 13. Explanation BASED LEARNING Known as Explanation Based Generalization. Form of Deductive Generalization. INPUTS USED BY EBL PROGRAM: Training example. Goal concept. Operational criteria Domain Theory
- 14. STEPS OF EBL PROGRAM--- Using Domain Theory (PRUNING ) Training Example Goal Concept Explanation Generalization
- 15. EXAMPLE : Training Example Owner(Object,Aakash) & Color(Object,Brown)& Is(Object,Light) & Has(Object,Coffee) & HasPart(Object,Concavity) & ...... PRUNING Domain Knowledge Is(X,Light)& HasPart(X,Y) & Is(Y,Handle) -> Liftable (X) HasPart(X,Y) & Is a(Y,Bottom) & Is(Y,Flat) -> Stable (X) HasPart(X,Y) & Is a(Y,Concavity) & Is(Y,UpwordPointing)->OpenVessel (X) Goal Concept [ CUP ] X is a Cup iff X is Stable , Liftable, OpenVessel Operationality Criterion Liftable --- Light Stable ---- Flat Open Vessel --- Concavity [ CUP ]
- 16. THANK YOU ANY QUESTIONS ?????...... PRESENTED BY: Amit Kumar Pandey SUBMITTED TO: Mr. NEERAJ KHARYA ( Lecturer MCA Dept )