Chapter 8 -FIND-S Algorithm
Chapter Outline
- Chapter Outline
- Quick Recap
- Research Cycle
- Representing Training Examples and Hypothesis for FIND-S Machine Learning Algorithm
- FIND-S Algorithm – Machine Learning Cycle
- Chapter Summary
Quick Recap
- Quick Recap – Concept Learning and Hypothesis Representation
- Learning is a Searching Problem
- Research is defined as a systematic investigation of sources and materials, to develop solutions for Real-world Problems to improve the quality (peace and happiness) of human life
- A Researcher is a man of character and (s)he should safeguard his character
- The word Research is made up of
- Re and
- search
- Therefore, in Research, we mainly refine the solution(s) proposed for a Real-world Problem
- The main steps of a Research Cycle are as follows
- Step 1: Identify the Real-world Problem
- Step 2: Propose Solution (called Solution 1) to solve the Real-world Problem
- Step 3: List down Strengths and Weaknesses of Solution 1
- Step 4: Propose Solution (called Solution 2) to
- further strengthen the Strengths of Solution 1
- overcome limitations of Solution 01
- Step 5: List down Strengths and Weaknesses of Solution 2
- Step 6: Propose Solution (called Solution 3) to
- further strengthen the Strengths of Solution 2
- overcome limitations of Solution 02
- Step 7: Continue this cycle till the Day of Judgment 😊
- Considering FIND-S Algorithm
- Input to Learner (FIND-S Algorithm)
- Set of Training Examples (D)
- Set of Functions / Hypothesis (H)
- Output by Learner (FIND-S Algorithm)
- A Hypothesis (h) from H which best fits the Training Examples (D)
- Note that h is an approximation of Target Function
- Inductive Bias Is the set of assumptions needed in addition to Training Examples to justify Deductively Learner’s Classification
- FIND-S Algorithm – Summary
- Representation of Example
- Attribute-Value Pair
- Representation of Hypothesis (h)
- Conjunction (AND) of Constraints on Attributes
- Searching Strategy
- I am not clear about this. Please drop me an email if you know. Jazak Allah Khair
- Training Regime
- Incremental Method
- Inductive Bias of FIND-S Algorithm
- Training Data is error-free
- Target Function / Concept is present in the Hypothesis Space (H)
- Strengths
- Returns a Model (h), which can be used to make predictions on unseen data
- Weaknesses
- Only works on error-free Data
- However, Real-world Data is noisy
- Works on assumption that Target Function is present in the Hypothesis Space (H)
- However, we may / may not find the Target Function in the Hypothesis Space (H) and this may / may not be known
- Only returns one hypothesis which best fits the Training Data
- However, there can be multiple hypothesis, which best fit the Training Data
Research Cycle
- Learning
- Research
- Definition
- A systematic investigation of sources and materials, to develop solutions for Real-world Problems to improve the quality (peace and happiness) of human life
- Purpose
- Develop solutions to serve the humanity for the sake of Allah and ultimately recognize Allah
- Importance
- In every era of human life, there will be various problems and it is important to develop efficient solutions to those problem
- Solutions to emerging problems are not possible without Research
- Applications
- Research helps us to understand
- How to get معرفت of الله (Who created us)?
- How to become a balanced and characterful human being?
- How to develop solutions to Real-world Problems to serve the humanity for the sake of Allah?
- Researcher
- Definition
- A Researcher is a man of character and (s)he should safeguard his character
- Main Characteristics of a Researcher
- عاجزی(Humbleness)
- سادگی (Simplicity)
- Only develops solutions / products which are beneficial for humanity
- Only develops solutions / products which help human beings to become a balanced and characterful personality
- Understanding the Word Research
- The word Research is made up of
- Re and
- search
- Therefore, in Research, we mainly refine the solution(s) proposed for a Real-world Problem
- Research Cycle
- The main steps of a Research Cycle are as follows
- Step 1: Identify the Real-world Problem
- Step 2: Propose Solution (called Solution 01) to solve the Real-world Problem
- Step 3: List down Strengths and Weaknesses of Solution 01
- Step 4: Propose Solution (called Solution 02) to
- further strengthen the Strengths of Solution 01
- overcome limitations of Solution 01
- Step 5: List down Strengths and Weaknesses of Solution 02
- Step 6: Propose Solution (called Solution 03) to
- further strengthen the Strengths of Solution 02
- overcome limitations of Solution 02
- Step 7: Continue this cycle till the Day of Judgment 😊
- Example 01 - Research Cycle
- Real-world Problem
- It gets dark after sunset and we cannot do anything
- Proposed Solution 01
- Produce sparks by rubbing two stones and burn woods to get fire
- Strengths
- We can get light by burning woods
- Limitations
- It requires a lot of effort to rub stones to get sparks
- It is not possible to move burning woods from one place to another
- Proposed Solution 02
- Match sticks were made to get spark to burn woods
- Strengths
- It overcomes the problem of Rubbing Stones to get sparks
- Limitations
- The duration and quantity of light obtained by Match Sticks is very short
- However, it is much better than Rubbing Stones Solution 😊
- It is not possible to move burning woods from one place to another
- Proposed Solution 03
- Candle
- Strengths
- The duration and quantity of light obtained is much better than Match Sticks
- We can easily take a Candle from one place to another and use its light to do tasks in the dark compared to Burning Woods
- Limitations
- Light generated by burning a Candle is
- For short time
- In small quantity
- Proposed Solution 04
- Bulb
- Strengths
- The duration and quantity of light generated by a Bulb is much better than a Candle
- Limitations
- Bulb fuses very quickly
- We need large power houses to generate current to enlighten a Bulb
- Proposed Solution 05
- Solutions to solve above problems are continuously proposed since the invention of the Bulb and will continue till the Day of Judgment In sha Allah😊
Example 02 – Research Cycle
- Real-world Problem
- How we can develop Intelligent Programs (Machine Learning Models) to perform various useful tasks
- Proposed Solution 01
- FIND-S Machine Learning Algorithm
- Strengths
- We can develop an Intelligent Programs using FIND-S Machine Learning Algorithm
- Weaknesses
- FIND-S Machine Learning Algorithm
- is a very simple algorithm and cannot handle complex Machine Learning Problems
- cannot learn from noisy Data and Real-world Data is noisy
- returns only one possible solution to a problem, but there can be many
- Proposed Solution 02
- Version Space Machine Learning Algorithms
- Strengths
- returns all possible solution to a problem
- Weaknesses
- Version Space Machine Learning Algorithm
- is a very simple algorithm and cannot handle complex Machine Learning Problems
- cannot learn from noisy Data and Real-world Data is noisy
- Proposed Solution 03
- Decision Tree Learning Algorithms
- Strengths
- Can handle complex Machine Learning Problems
- Can learn from noisy Data
- Weaknesses
- Cannot handle very complex Numeric Representations of Data
- Cannot handle very complex Machine Learning Problems
- Proposed Solution 04
- Artificial Neural Networks
- Strengths
- Can handle noisy Data
- Can handle very complex Machine Learning Problems
- Can handle very complex Numeric Representation of Data
- Weaknesses
- Artificial Neural Networks (particularly Deep Learning Models) require high computational resources
- Artificial Neural Networks (particularly Deep Learning Models) require high Training Time
- Proposed Solution 05
- Research community is continuously doing Research to solve the above-mentioned problems and it will continue till the Day of Judgment In sha Allah😊
TODO and Your Turn
TODO Task 1
- Task 1
- The first Light Bulb was made in 1879. A picture of how Light Bulb looked like in 1879 is given below
- Note
- Your answer should be
- Well Justified
- Questions
- Write Input and Output for a Light Bulb?
- Write a short Biography of Thomas Edison (one Page) and list down the lessons you learned from it?
- How many attempts Thomas Edison made before he succeeded in making a Light Bulb? What lesson you learned from his attempts?
- Execute the Research Cycle for Light Bulb from 1879 to 2020. Also, put Pictures of Light Bulbs (with Year) whenever there was a major breakthrough in the technology?
Your Turn Task 1
- Task 1
- Select a Task (similar to Light Bulb) and answer the questions given below.
- Questions
- Write Input and Output for the selected Task?
- Write a short Biography of the inventor of the Task and list down the lessons you learned from it?
- Execute the Research Cycle for selected Task.
- If possible, also put Pictures of selected Task (with Year) whenever there was a major breakthrough in the technology?
Representing Training Examples and Hypothesis for FIND-S Machine Learning Algorithm
- Learning Input-Output Functions – General Settings
- Input to Learner
- Set of Training Examples (D)
- Set of Functions / Hypothesis (H)
- Output by Learner
- A Hypothesis (h) from H which best fits the Training Examples (D)
- Note that h is an approxicmation of Target Functino f
- In this Chapter
- Learner is FIND-S Machine Learning Algorithm 😊
- Chapter Focus
- In sha Allah (انشاء اللہ), in this Chapter, I will take the Gender Identification Problem and try to explain three main things
- Representation of Training Examples (D)
- How to represent Training Examples (D) in a Format which FIND-S Algorithm can understand and learn from them?
- Representation of Hypothesis (h)
- How to represent Hypothesis (h) in a Format which FIND-S Algorithm can understand?
- Searching Strategy
- What searching strategy is used by FIND-S Algorithm to find a h from H, which best fits the Training Examples (D)?
- Gender Identification Problem
- Machine Lerning Problem
- Gender Identification
- Input
- Human
- Output
- Gender of a Human
- Task
- Given a Human (Input), predict the Gender of the Human (Output)
- Treated as
- Learning Input-Output Function
- i.e., Learn from Input to predict Output
- Representation of Examples
- Example = Input + Output
- Input = Human
- Output = Gender
- Representatino of Input and Output
- Attribute-Value Pair
- Representation of Input
- Input is represented as a
- Set of 6 Input Attributes
- Height
- Weight
- HairLength
- HeadCovered
- WearingChain
- ShirtSleeves
- Representation of Output
- Input is represented as a
- Set of 1 Output Attribute
- Gender
- Note
- Yes means Female and No means Male
- Computing Size of Instance Space (X)
- Computing Size of Instance Space (X)
- |X| = No. of Values of Height Input Attribute x No. of Values of Weight Input Attribute x No. of Values of HairLength Input Attribute x No. of Values of HeadCovered Input Attribute x No. of Values of WearingChain Input Attribute x No. of Values of ShirtSleeves Input Attribute
- |X| = 3*2*2*2*2*2 = 96
- Sample Data
- We obtained a Sample Data of 6 examples
- Representation of Hypothesis (h)
- We represent a Hypothesis (h) as
- Conjunction (AND) of Constraints on Input Attributes
- Each constraint can be:
- No value allowed (null hypothesis Ø): e.g. Height = Ø
- A specific value : e.g. Height = Short
- A don’t care value (any of possible values): e.g. Height = ?
- Representation of Hypothesis (h) Cont...
- Most Specific Hypothesis (h)
< Height Weight HairLength HeadCovered WearingChain ShirtSleeves >
< ∅ ∅ ∅ ∅ ∅ ∅ >
- Most General Hypothesis (h)
< Height Weight HairLength HeadCovered WearingChain ShirtSleeves >
< ? ? ? ? ? ? >
- Another Hypothesis (h)
< Height Weight HairLength HeadCovered WearingChain ShirtSleeves >
< Normal Light ? ? No ? >
- Importance Note
- The order of Input Attributes must be exactly same in
- Training Example (d) and Hypotheis (h)
- Computing Size of Concept Spane (C) and Hypothesis Space (H)
- Size of Instance Space (X)
- |X| = 96
- Size of Concept Space (C)
- |C| = 2|X| = 2|96| = 79,228,162,514,264,337,593,543,950,336
- Size of Hypothesis Space (H) (Syntactically Distinct Hypothesis)
- |H| = 5*4*4*4*4*4 = 5120
- Size of Hypothesis Space (H) (Semantically Distinct Hypothesis)
- |H| = 1+4*3*3*3*3*3 = 973
FIND-S Algorithm - Machine Learning Cycle
- Machine Learning Cycle
- Four phases of a Machine Learning Cycle are
- Training Phase
- Build the Model using Training Data
- Testing Phase
- Evaluate the performance of Model using Testing Data
- Application Phase
- Deploy the Model in Real-world, to make prediction on Real-time Unseen Data
- Feedback Phase
- Take Feedback form the Users and Domain Experts to improve the Model
- Split the Sample Data
- We split the Sample Data using Random Split Approach into
- Training Data – 2 / 3 of Sample Data
- Testing Data – 1 / 3 of Sample Data
- Sample Data
- Training Data
- Testing Data
- Note
- Sample Data is balanced
- 3 Postivie Instances (Female)
- 3 Negative Instances (Male)
- After splitting Sample Data using Random Split Approach
- Training Data is unbalanced
- 3 Postivie Instances (Female)
- 1 Negative Instances (Male)
- Testing Data is unbalanced
- 0 Postivie Instances (Female)
- 2 Negative Instances (Male)
- Sample Data – Vector Representation
- Vector Representation of Examples
x1= <Short, Light, Short, Yes, Yes, Half> +
x2= <Short, Light, Long, Yes, Yes, Half> +
x3 = <Tall, Heavy, Long, Yes, Yes, Full> –
x4 = <Short, Light, Long, Yes, No, Full> +
x5 = <Short, Light, Short, Yes, Yes, Half> –
x6 = <Tall, Light, Short, No, Yes, Full> –
- Training Data – Vector Representation
- Vector Representation of Training Examples
x1= <Short, Light, Short, Yes, Yes, Half> +
x2= <Short, Light, Long, Yes, Yes, Half> +
x3 = <Tall, Heavy, Long, Yes, Yes, Full> –
x4 = <Short, Light, Long, Yes, No, Full> +
- Testing Data – Vector Representation
- Vector Representation of Test Examples
x5 = <Short, Light, Short, Yes, Yes, Half> –
x6 = <Tall, Light, Short, No, Yes, Full> –
- Find-S Algorithm (or Learner)
- Specific to General Constraints
- We have three constraints on our Attributes
- No value allowed (Ø)
- Most Specific Constraint
- A set of specific values (for e.g. Short, Normal and Tall for Height Attribute)
- Note that specific value is next more generic constraint then No value allowed (Ø)
- A don’t care value (?)
- Note that ? is next more generic constraint then specific value
- Training Phase
- In the Training Phase, the FIND-S Algorithm will
- Seearch H to find out a h, which best fits the Training Data
- Best Fit means
- h correctly classifies Positive and Negative instances in the Training Data
- Correct Classification
- Postive instance is classified as Positive
- Negative instance is classified as Negative
- Inorrect Classification
- Negative instance is classified as Positive
- Postive instance is classified as Negative
- Initialize h to the Most Specific Hypothesis in H
- h0 = <∅ , ∅ ,∅ ,∅ ,∅ , ∅>
- For each positive training instance x
- For each attribute constraint ai in h
- If the constraint ai in h is satisfied by x
- then do nothing
- else replace ai in h by the next more general constraint that is satisfied by x
- First Training Example
- x1 = <Short, Light, Short, Yes, Yes, Half> +
- Let’s see if attribute constraints in h0 satisfy x1 or not?
- As we can see that attribute constraints in h0 do not satisfy x1
- Therefore, x1 is incorrectly classified as Negative
- To satisfy x1, we will need to replace attribute constaints in h0 by the next more general constraint that is satisfied by x1
- h0 = <∅ , ∅ ,∅ ,∅ ,∅ , ∅> will become
- h1 = <Short, Light, Short, Yes, Yes, Half>
- First Training Example
- x1 = <Short, Light, Short, Yes, Yes, Half> +
- Let’s see if attribute constraints in h1 satisfy x1 or not?
- As we can see that attribute constraints in h1 satisfy x1
- Therefore, x1 is correctly classified as Positive
- Second Training Example
- x2 = <Short, Light, Long, Yes, Yes, Half> +
- Let’s see if attribute constraints in h1 satisfy x2 or not?
- As we can see that attribute constraints in h1 do not satisfy x2
- Therefore, x2 is incorrectly classified as Negative
- To satisfy x2, we will need to replace attribute constaints in h1 by the next more general constraint that is satisfied by x2
- h1 = <Short, Light, Short, Yes, Yes, Half> will become
- h2 = <Short, Light, ?, Yes, Yes, Half>
- Second Training Example
- x2 = <Short, Light, Long, Yes, Yes, Half> +
- Let’s see if attribute constraints in h2 satisfy x2 or not?
- As we can see that attribute constraints in h1 satisfy x2
- Therefore, x2 is correctly classified as Positive
- Note
- Learner (FIND-S Algorithm) has observed two Training examples up till now and our hypothesis is as follows
- h2 = <Short, Light, ?, Yes, Yes, Half>
- Let’s see if h2 must best fit the observed Trailing Example i.e. x1 and x2 or not?
- h2 correctly classifies
- x1 as Positive
- x2 as Positive
- To conclude, h2 best fits’ the first two observed Trailing Example i.e. x1 and x2
- Third Training Example
- x3 = <Tall, Heavy, Long, Yes, Yes, Full> –
- Note that 3rd Training Example is Negative and
- FIND-S only operates on Positive Training Examples
- Therefore, there will be on change in h2 and h3
- h2 = <Short, Light, ?, Yes, Yes, Half> will become
- h3 = <Short, Light, ?, Yes, Yes, Half>
- Interesting h3 correctly classifies x3 as Negative
- Third Training Example
- x3 = <Tall, Heavy, Long, Yes, Yes, Full> –
- h3 correctly classifies
- x1 as Positive
- x2 as Positive
- x3 as Negative
- Thus, h3 best fits the three Training Examples observed up till now
- Fourth Training Example
- x4 = <Short, Light, Long, Yes, No, Full> +
- Let’s see if attribute constraints in h3 satisfy x4 or not?
- As we can see that attribute constraints in h3 do not satisfy x4
- Therefore, x4 is incorrectly classified as Negative
- To satisfy x4, we will need to replace attribute constaints in h3 by the next more general constraint that is satisfied by x4
- h3 = <Short, Light, ?, Yes, , Half>
- h4 = <Short, Light, ?, Yes, ?, ?>
- Fourth Training Example
- x4 = <Short, Light, Long, Yes, No, Full> +
- Let’s see if attribute constraints in h4 satisfy x4 or not?
- As we can see that attribute constraints in h4 satisfy x4
- Therefore, x4 is correctly classified as Positive
- h4 correctly classifies
- x1 as Positive
- x2 as Positive
- x3 as Negative
- x4 as Positive
- Thus, h4 best fits the three Training Examples observed up till now
- Note
- There were total 4 Training Examples and we have observed all of them
FIND-S Algorithm
- After observing all the Training Example, the FIND-S Algorithm will
- Output hypothesis h
- The h returned by FIND-S Algorithm is
- h = <Short, Light, ?, Yes, ?, ?>
- Note that h is an approximation of the Target Function f
- Recall
- Training Data
- x1 = <Short, Light, Short, Yes, Yes, Half> +
- x2 = <Short, Light, Long, Yes, Yes, Half> +
- x3 = <Tall, Heavy, Long, Yes, Yes, Full> –
- x4 = <Short, Light, Long, Yes, No, Full>
- Model
- h = <Short, Light, ?, Yes, ?, ?>
- Model – in the form of Rules
- In the next phase i.e. Testing Phase, we will
- Evaluate the performance of the Model
- Testing Phase
- Question
- How good Model has learned?
- Answer
- Evaluate the performance of Model on unseen data (or Testing Data)
- Evolution Measures
- Evaluation will be carried out using
- Error measure
- Error
- Definition
- Error is defined as the proportion of incorrectly classified Test instances
- Formula
- Note
- Accuracy = 1 – Error
- Evaluate Model
- Applying Model on x5
- Prediction returned by Model
- x5 is predicted Positive (Incorrectly Classified Instance)
- Applying Model on x6
- Prediction returned by Model
- x6 is predicted Negative (Correctly Classified Instance)
- Evaluate Model, Cont…
- Application Phase
- We assume that our Model
- performed well on large Test Data and can be deployed in Real-world
- Model is deployed in the Real-world and now we can make
- predictions on Real-time Data
- Steps – Making Predictions on Real-time Data
- Step 1: Take Input from User
- Step 2: Convert User Input into Feature Vector
- Exactly same as Feature Vectors of Training and Testing Data
- Step 3: Apply Model on the Feature Vector
- Step 4: Return Prediction to the User
- Example - Making Predictions on Real-time Data
- Step 1: Take input from User
- Step 2: Convert User Input into Feature Vector
- Note that order of Attributes must be exactly same as that of Training and Testing Examples
Step 3: Apply Model on the Feaure Vector
- Step 4: Return Prediction to the User
- Positive
- Note
- You can take Input from user, apply Model and return predictions as many times as you like 😊
- Feedback Phase
- Only Allah is Perfect 😊
- Take Feedback on your deployed Model from
- Domain Experts and
- Users
- Improve your Model based on Feedback 😊
- Inductive Bias - FIND-S Algorithm
- Inductive Bias
- Inductive Bias Is the set of assumptions needed in addition to Training Examples to justify Deductively Learner’s Classification
- Inductive Bias of FIND-S Algorithm
- Training Data is error free
- Target Function / Concept is present in the Hypothesis Space (H)
- Training Regime- FIND-S Algorithm
- Incremental Method
- See Chapter 3 – Basics of Machine Learning
- Strengths and Weakness - FIND-S Algorithm
- Strengths
- Returns a Model (h), which can be used to make predictions on unseen data
- Weaknesses
- Only works on error free Data
- However, Real-world Data is noisy
- Works on assumption that Target Function is present in the Hypothesis Space (H)
- However, we may / may not find the Target Function in the Hypothesis Space (H) and this may / may not be known
- Only returns one hypothesis which best fits the Training Data
- However, there can be multiple hypothesis, which best fit the Training Data
TODO and Your Turn
TODO Task 2
- Task 1
- Consider the Titanic Dataset with the following Attributes
- Gender: Male, Female
- Ticket Class: Upper, Middle, Lower
- Parent/Child Abroad: Zero, One, Two, Three
- Embarked: Cherbourg, Queenstown, Southampton
- Survival: No, Yes
- We obtained the following Sample Data
- Sample Data was split into Training and Testing Data in a
- Train-Test Split Ratio of 67%-33%
- Training Data
- Testing Data
- Note
- Consider FIND-S Algorithm when answering questions given below
- Your answer should be
- Well Justified
- Questions
- Write down the Input and Output for the above Machine Learning Problem?
- How Training Example is represented?
- How Hypothesis (h) should be represented?
- Calculate Size of Instance Space, Concept Space, Syntactically Distinct Hypothesis and Semantically Distinct Hypothesis?
- Executer the Machine Learning Cycle?
- Write down your observations that you observed during the execution of Machine Learning Cycle?
Your Turn Task 2
- Task 1
- Select a Machine Learning Problem (similar to: Titanic – Machine Learning form Disaster) and answer the questions given below.
- Note
- Consider FIND-S Algorithm in answering all the questions.
- Questions
- Write down the Input and Output for the selected Machine Learning Problem?
- How Training Example is represented?
- How Hypothesis (h) should be represented?
- Calculate Size of Instance Space, Concept Space, Syntactically Distinct Hypothesis and Semantically Distinct Hypothesis?
- Executer the Machine Learning Cycle?
- Write down your observations that you observed during the execution of Machine Learning Cycle?
Chapter Summary
- Chapter Summary
- Learning is a Searching Problem
- Research is defined as a systematic investigation of sources and materials , to develop solutions for Real-world Problems to improve the quality (peace and happiness) of human life
- A Researcher is a man of character and (s)he should safeguard his character
- The word Research is made up of
- Re and
- search
- Therefore, in Research, we mainly refine the solution(s) proposed for a Real-world Problem
- The main steps of a Research Cycle are as follows
- Step 1: Identify the Real-world Problem
- Step 2: Propose Solution (called Solution 01) to solve the Real-world Problem
- Step 3: List down Strengths and Weaknesses of Solution 01
- Step 4: Propose Solution (called Solution 02) to
- further strengthen the Strengths of Solution 01
- overcome limitations of Solution 01
- Step 5: List down Strengths and Weaknesses of Solution 02
- Step 6: Propose Solution (called Solution 03) to
- further strengthen the Strengths of Solution 02
- overcome limitations of Solution 02
- Step 7: Continue this cycle till the Day of Judgment 😊
- Considering FIND-S Algorithm
- Input to Learner (FIND-S Algorithm)
- Set of Training Examples (D)
- Set of Functions / Hypothesis (H)
- Output by Learner (FIND-S Algorithm)
- A Hypothesis (h) from H which best fits the Training Examples (D)
- Note that h is an approximation of Target Function
- Inductive Bias Is the set of assumptions needed in addition to Training Examples to justify Deductively Learner’s Classification
- FIND-S Algorithm – Summary
- Representation of Example
- Attribute-Value Pair
- Representation of Hypothesis (h)
- Conjunction (AND) of Constraints on Attributes
- Searching Strategy
- I am not clear about this. Please drop me an email if you know. Jazak Allah Khair
- Training Regime
- Incremental Method
- Inductive Bias of FIND-S Algorithm
- Training Data is error-free
- Target Function / Concept is present in the Hypothesis Space (H)
- Strengths
- Returns a Model (h), which can be used to make predictions on unseen data
- Weaknesses
- Only works on error-free Data
- However, Real-world Data is noisy
- Works on assumption that Target Function is present in the Hypothesis Space (H)
- However, we may / may not find the Target Function in the Hypothesis Space (H) and this may / may not be known
- Only returns one hypothesis which best fits the Training Data
- However, there can be multiple hypothesis , which best fit the Training Data
In Next Chapter
- In Next Chapter
- In Sha Allah, in next Chapter, I will present
- Version Space Algorithm