SlideShare une entreprise Scribd logo
1  sur  20
Curse of Dimensionality

                            By,
                 Nikhil Sharma
What is it?
   In applied mathematics, curse of
    dimensionality (a term coined by Richard E.
    Bellman), also known as the Hughes
    effect (named after Gordon F. Hughes),refers to
    the problem caused by the exponential increase
    in volume associated with adding extra
    dimensions to a mathematical space.
Problems
•   High Dimensional Data is difficult to work with
    since:
      Adding more features can increase the noise, and
       hence the error
      There usually aren’t enough observations to get
       good estimates
•   This causes:
      Increase in running time
      Overfitting
      Number of Samples Required
An Example
Representation of 10% sample probability space
    (i) 2-D                       (ii)3-D




The Number of Points Would Need to Increase Exponentially
              to Maintain a Given Accuracy.
 10n samples would be required for a n-dimension problem.
Curse of Dimensionality:
               Complexity
   Complexity (running time) increases with
    dimension d
    A lot of methods have at least
                                        O(nd 2 )
    complexity, where n is the number of samples
   So as d becomes large, O(nd2) complexity may
    be too costly
Curse of Dimensionality:
               Overfitting
   Paradox: If n < d2 we are better off assuming
    that features are uncorrelated, even if we know
    this assumption is wrong
   We are likely to avoid overfitting because we fit a
    model with less parameters:
Curse of Dimensionality: Number
               of Samples
 Suppose we want to use the nearest neighbor
  approach with k = 1 (1NN)
 This feature is not discriminative, i.e. it does not
 separate the classes well
 Suppose we start with only one feature




 We decide to use 2 features. For the 1NN method to
  work well, need a lot of samples, i.e. samples have to
  be dense
 To maintain the same density as in 1D (9 samples per
  unit length), how many samples do we need?
Curse of Dimensionality: Number
           of Samples
   We need 92 samples to maintain the same
    density as in 1D
Curse of Dimensionality: Number
           of Samples
   Of course, when we go from 1 feature to 2, no
    one gives us more samples, we still have 9




   This is way too sparse for 1NN to work well
Curse of Dimensionality: Number
               of Samples
   Things go from bad to worse if we decide to use 3
    features:




   If 9 was dense enough in 1D, in 3D we need 93=729
    samples!
Curse of Dimensionality: Number
           of Samples
   In general, if n samples is dense enough in 1D

   Then in d dimensions we need n d samples!
   And n d grows really really fast as a function of d
   Common pitfall:
       If we can’t solve a problem with a few features, adding
         more features seems like a good idea
       However the number of samples usually stays the same
       The method with more features is likely to perform worse
        instead of expected better
Curse of Dimensionality: Number
           of Samples
   For a fixed number of samples, as we add
    features, the graph of classification error:




   Thus for each fixed sample size n, there is the
    optimal number of features to use
The Curse of Dimensionality
 We should try to avoid creating lot of features
 Often no choice, problem starts with many features
 Example: Face Detection
     One sample point is k by m array of pixels




       Feature extraction is not trivial, usually every
       pixel is taken as a feature
        Typical dimension is 20 by 20 = 400
        Suppose 10 samples are dense enough for 1 dimension.
        Need only 10400samples
The Curse of Dimensionality
   Face Detection, dimension of one sample point is km




   The fact that we set up the problem with km dimensions
    (features) does not mean it is really a km-dimensional
    problem
    Most likely we are not setting the problem up with the right
    features
    If we used better features, we are likely need much less than
    km-dimensions
    Space of all k by m images has km dimensions
    Space of all k by m faces must be much smaller, since faces
    form a tiny fraction of all possible images
Dimensionality Reduction
   We want to reduce the number of
    dimensions because:
      Efficiency.
      Measurement costs.
      Storage costs.
      Computation costs.
      Classification performance.
      Ease of interpretation/modeling.
Principal Components
   The idea is to project onto the subspace which accounts for
    most of the variance.

   This is accomplished by projecting onto the eigenvectors of
    the covariance matrix associated with the largest
    eigenvalues.

   This is generally not the projection best suited for
    classification.

   It can be a useful method for providing a first-cut reduction
    in dimension from a high dimensional space
Feature Combination as a method
    to reduce Dimensionality
   High dimensionality is challenging and redundant
    It is natural to try to reduce dimensionality
    Reduce dimensionality by feature combination: combine
    old features x to create new features y




   For Example,



   Ideally, the new vector y should retain from x all
    information important for classification
Principle Component Analysis
                  (PCA)
   Main idea: seek most accurate data representation in a
    lower dimensional space
   Example in 2-D
        ◦ Project data to 1-D subspace (a line) which minimize the
          projection error




   Notice that the the good line to use for projection lies in
    the direction of largest variance
PCA: Approximation of Elliptical
         Cloud in 3D
Thank You!
   The End.

Contenu connexe

Tendances

Data Mining:Concepts and Techniques, Chapter 8. Classification: Basic Concepts
Data Mining:Concepts and Techniques, Chapter 8. Classification: Basic ConceptsData Mining:Concepts and Techniques, Chapter 8. Classification: Basic Concepts
Data Mining:Concepts and Techniques, Chapter 8. Classification: Basic ConceptsSalah Amean
 
K-Nearest Neighbor Classifier
K-Nearest Neighbor ClassifierK-Nearest Neighbor Classifier
K-Nearest Neighbor ClassifierNeha Kulkarni
 
07 dimensionality reduction
07 dimensionality reduction07 dimensionality reduction
07 dimensionality reductionMarco Quartulli
 
Machine Learning With Logistic Regression
Machine Learning  With Logistic RegressionMachine Learning  With Logistic Regression
Machine Learning With Logistic RegressionKnoldus Inc.
 
2.2 decision tree
2.2 decision tree2.2 decision tree
2.2 decision treeKrish_ver2
 
Understanding Bagging and Boosting
Understanding Bagging and BoostingUnderstanding Bagging and Boosting
Understanding Bagging and BoostingMohit Rajput
 
Machine Learning and Real-World Applications
Machine Learning and Real-World ApplicationsMachine Learning and Real-World Applications
Machine Learning and Real-World ApplicationsMachinePulse
 
Linear regression in machine learning
Linear regression in machine learningLinear regression in machine learning
Linear regression in machine learningShajun Nisha
 
Introduction to Autoencoders
Introduction to AutoencodersIntroduction to Autoencoders
Introduction to AutoencodersYan Xu
 
Autoencoder
AutoencoderAutoencoder
AutoencoderHARISH R
 
Machine Learning with Decision trees
Machine Learning with Decision treesMachine Learning with Decision trees
Machine Learning with Decision treesKnoldus Inc.
 
Machine learning with ADA Boost
Machine learning with ADA BoostMachine learning with ADA Boost
Machine learning with ADA BoostAman Patel
 
Transfer learning-presentation
Transfer learning-presentationTransfer learning-presentation
Transfer learning-presentationBushra Jbawi
 
Machine learning
Machine learningMachine learning
Machine learningRohit Kumar
 
Decision Tree - C4.5&CART
Decision Tree - C4.5&CARTDecision Tree - C4.5&CART
Decision Tree - C4.5&CARTXueping Peng
 

Tendances (20)

Data Mining:Concepts and Techniques, Chapter 8. Classification: Basic Concepts
Data Mining:Concepts and Techniques, Chapter 8. Classification: Basic ConceptsData Mining:Concepts and Techniques, Chapter 8. Classification: Basic Concepts
Data Mining:Concepts and Techniques, Chapter 8. Classification: Basic Concepts
 
K-Nearest Neighbor Classifier
K-Nearest Neighbor ClassifierK-Nearest Neighbor Classifier
K-Nearest Neighbor Classifier
 
Support Vector Machines ( SVM )
Support Vector Machines ( SVM ) Support Vector Machines ( SVM )
Support Vector Machines ( SVM )
 
07 dimensionality reduction
07 dimensionality reduction07 dimensionality reduction
07 dimensionality reduction
 
Machine Learning With Logistic Regression
Machine Learning  With Logistic RegressionMachine Learning  With Logistic Regression
Machine Learning With Logistic Regression
 
Supervised learning
  Supervised learning  Supervised learning
Supervised learning
 
2.2 decision tree
2.2 decision tree2.2 decision tree
2.2 decision tree
 
Understanding Bagging and Boosting
Understanding Bagging and BoostingUnderstanding Bagging and Boosting
Understanding Bagging and Boosting
 
Machine Learning and Real-World Applications
Machine Learning and Real-World ApplicationsMachine Learning and Real-World Applications
Machine Learning and Real-World Applications
 
Machine learning
Machine learningMachine learning
Machine learning
 
Linear regression in machine learning
Linear regression in machine learningLinear regression in machine learning
Linear regression in machine learning
 
Introduction to Autoencoders
Introduction to AutoencodersIntroduction to Autoencoders
Introduction to Autoencoders
 
Autoencoder
AutoencoderAutoencoder
Autoencoder
 
Machine Learning with Decision trees
Machine Learning with Decision treesMachine Learning with Decision trees
Machine Learning with Decision trees
 
Machine learning with ADA Boost
Machine learning with ADA BoostMachine learning with ADA Boost
Machine learning with ADA Boost
 
Naive Bayes
Naive BayesNaive Bayes
Naive Bayes
 
Transfer learning-presentation
Transfer learning-presentationTransfer learning-presentation
Transfer learning-presentation
 
Confusion Matrix Explained
Confusion Matrix ExplainedConfusion Matrix Explained
Confusion Matrix Explained
 
Machine learning
Machine learningMachine learning
Machine learning
 
Decision Tree - C4.5&CART
Decision Tree - C4.5&CARTDecision Tree - C4.5&CART
Decision Tree - C4.5&CART
 

Similaire à Curse of dimensionality

17- Kernels and Clustering.pptx
17- Kernels and Clustering.pptx17- Kernels and Clustering.pptx
17- Kernels and Clustering.pptxssuser2023c6
 
An Efficient Method of Partitioning High Volumes of Multidimensional Data for...
An Efficient Method of Partitioning High Volumes of Multidimensional Data for...An Efficient Method of Partitioning High Volumes of Multidimensional Data for...
An Efficient Method of Partitioning High Volumes of Multidimensional Data for...IJERA Editor
 
Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tu...
Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tu...Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tu...
Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tu...San Kim
 
Machine Learning lecture6(regularization)
Machine Learning lecture6(regularization)Machine Learning lecture6(regularization)
Machine Learning lecture6(regularization)cairo university
 
Dimensionality Reduction
Dimensionality ReductionDimensionality Reduction
Dimensionality ReductionSaad Elbeleidy
 
Density maximization for improving graph matching with its applications
Density maximization for improving graph matching with its applicationsDensity maximization for improving graph matching with its applications
Density maximization for improving graph matching with its applicationsI3E Technologies
 
Dimensionality Reduction | Machine Learning | CloudxLab
Dimensionality Reduction | Machine Learning | CloudxLabDimensionality Reduction | Machine Learning | CloudxLab
Dimensionality Reduction | Machine Learning | CloudxLabCloudxLab
 
Machine learning using matlab.pdf
Machine learning using matlab.pdfMachine learning using matlab.pdf
Machine learning using matlab.pdfppvijith
 
17 large scale machine learning
17 large scale machine learning17 large scale machine learning
17 large scale machine learningTanmayVijay1
 
Web image annotation by diffusion maps manifold learning algorithm
Web image annotation by diffusion maps manifold learning algorithmWeb image annotation by diffusion maps manifold learning algorithm
Web image annotation by diffusion maps manifold learning algorithmijfcstjournal
 
Making BIG DATA smaller
Making BIG DATA smallerMaking BIG DATA smaller
Making BIG DATA smallerTony Tran
 
Hands-On Machine Learning with Scikit-Learn and TensorFlow - Chapter8
Hands-On Machine Learning with Scikit-Learn and TensorFlow - Chapter8Hands-On Machine Learning with Scikit-Learn and TensorFlow - Chapter8
Hands-On Machine Learning with Scikit-Learn and TensorFlow - Chapter8Hakky St
 
"Designing CNN Algorithms for Real-time Applications," a Presentation from Al...
"Designing CNN Algorithms for Real-time Applications," a Presentation from Al..."Designing CNN Algorithms for Real-time Applications," a Presentation from Al...
"Designing CNN Algorithms for Real-time Applications," a Presentation from Al...Edge AI and Vision Alliance
 

Similaire à Curse of dimensionality (20)

PCA.pptx
PCA.pptxPCA.pptx
PCA.pptx
 
17- Kernels and Clustering.pptx
17- Kernels and Clustering.pptx17- Kernels and Clustering.pptx
17- Kernels and Clustering.pptx
 
An Efficient Method of Partitioning High Volumes of Multidimensional Data for...
An Efficient Method of Partitioning High Volumes of Multidimensional Data for...An Efficient Method of Partitioning High Volumes of Multidimensional Data for...
An Efficient Method of Partitioning High Volumes of Multidimensional Data for...
 
Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tu...
Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tu...Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tu...
Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tu...
 
Machine Learning lecture6(regularization)
Machine Learning lecture6(regularization)Machine Learning lecture6(regularization)
Machine Learning lecture6(regularization)
 
Clique and sting
Clique and stingClique and sting
Clique and sting
 
Dimensionality Reduction
Dimensionality ReductionDimensionality Reduction
Dimensionality Reduction
 
Density maximization for improving graph matching with its applications
Density maximization for improving graph matching with its applicationsDensity maximization for improving graph matching with its applications
Density maximization for improving graph matching with its applications
 
Dimensionality Reduction | Machine Learning | CloudxLab
Dimensionality Reduction | Machine Learning | CloudxLabDimensionality Reduction | Machine Learning | CloudxLab
Dimensionality Reduction | Machine Learning | CloudxLab
 
Machine learning using matlab.pdf
Machine learning using matlab.pdfMachine learning using matlab.pdf
Machine learning using matlab.pdf
 
deep CNN vs conventional ML
deep CNN vs conventional MLdeep CNN vs conventional ML
deep CNN vs conventional ML
 
17 large scale machine learning
17 large scale machine learning17 large scale machine learning
17 large scale machine learning
 
Explore ml day 2
Explore ml day 2Explore ml day 2
Explore ml day 2
 
Web image annotation by diffusion maps manifold learning algorithm
Web image annotation by diffusion maps manifold learning algorithmWeb image annotation by diffusion maps manifold learning algorithm
Web image annotation by diffusion maps manifold learning algorithm
 
Making BIG DATA smaller
Making BIG DATA smallerMaking BIG DATA smaller
Making BIG DATA smaller
 
Hands-On Machine Learning with Scikit-Learn and TensorFlow - Chapter8
Hands-On Machine Learning with Scikit-Learn and TensorFlow - Chapter8Hands-On Machine Learning with Scikit-Learn and TensorFlow - Chapter8
Hands-On Machine Learning with Scikit-Learn and TensorFlow - Chapter8
 
04 numerical
04 numerical04 numerical
04 numerical
 
"Designing CNN Algorithms for Real-time Applications," a Presentation from Al...
"Designing CNN Algorithms for Real-time Applications," a Presentation from Al..."Designing CNN Algorithms for Real-time Applications," a Presentation from Al...
"Designing CNN Algorithms for Real-time Applications," a Presentation from Al...
 
lec10svm.ppt
lec10svm.pptlec10svm.ppt
lec10svm.ppt
 
Svm ms
Svm msSvm ms
Svm ms
 

Plus de Nikhil Sharma

Plus de Nikhil Sharma (6)

Digital Life
Digital LifeDigital Life
Digital Life
 
B tree short
B tree shortB tree short
B tree short
 
B tree long
B tree longB tree long
B tree long
 
India
IndiaIndia
India
 
Impurities in wastewater & problems caused by it
Impurities in wastewater & problems caused by itImpurities in wastewater & problems caused by it
Impurities in wastewater & problems caused by it
 
Asymptotic notations
Asymptotic notationsAsymptotic notations
Asymptotic notations
 

Dernier

Understanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdfUnderstanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdfUK Journal
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...apidays
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfEnterprise Knowledge
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsEnterprise Knowledge
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityPrincipled Technologies
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)Gabriella Davis
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024The Digital Insurer
 
Tech Trends Report 2024 Future Today Institute.pdf
Tech Trends Report 2024 Future Today Institute.pdfTech Trends Report 2024 Future Today Institute.pdf
Tech Trends Report 2024 Future Today Institute.pdfhans926745
 
Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024The Digital Insurer
 
Evaluating the top large language models.pdf
Evaluating the top large language models.pdfEvaluating the top large language models.pdf
Evaluating the top large language models.pdfChristopherTHyatt
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘RTylerCroy
 
presentation ICT roal in 21st century education
presentation ICT roal in 21st century educationpresentation ICT roal in 21st century education
presentation ICT roal in 21st century educationjfdjdjcjdnsjd
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?Antenna Manufacturer Coco
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slidevu2urc
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking MenDelhi Call girls
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking MenDelhi Call girls
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
CNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of ServiceCNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of Servicegiselly40
 

Dernier (20)

Understanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdfUnderstanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdf
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivity
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
 
Tech Trends Report 2024 Future Today Institute.pdf
Tech Trends Report 2024 Future Today Institute.pdfTech Trends Report 2024 Future Today Institute.pdf
Tech Trends Report 2024 Future Today Institute.pdf
 
Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024
 
Evaluating the top large language models.pdf
Evaluating the top large language models.pdfEvaluating the top large language models.pdf
Evaluating the top large language models.pdf
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘
 
presentation ICT roal in 21st century education
presentation ICT roal in 21st century educationpresentation ICT roal in 21st century education
presentation ICT roal in 21st century education
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
CNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of ServiceCNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of Service
 

Curse of dimensionality

  • 1. Curse of Dimensionality By, Nikhil Sharma
  • 2. What is it?  In applied mathematics, curse of dimensionality (a term coined by Richard E. Bellman), also known as the Hughes effect (named after Gordon F. Hughes),refers to the problem caused by the exponential increase in volume associated with adding extra dimensions to a mathematical space.
  • 3. Problems • High Dimensional Data is difficult to work with since: Adding more features can increase the noise, and hence the error There usually aren’t enough observations to get good estimates • This causes: Increase in running time Overfitting Number of Samples Required
  • 4. An Example Representation of 10% sample probability space (i) 2-D (ii)3-D The Number of Points Would Need to Increase Exponentially to Maintain a Given Accuracy. 10n samples would be required for a n-dimension problem.
  • 5. Curse of Dimensionality: Complexity  Complexity (running time) increases with dimension d  A lot of methods have at least O(nd 2 ) complexity, where n is the number of samples  So as d becomes large, O(nd2) complexity may be too costly
  • 6. Curse of Dimensionality: Overfitting  Paradox: If n < d2 we are better off assuming that features are uncorrelated, even if we know this assumption is wrong  We are likely to avoid overfitting because we fit a model with less parameters:
  • 7. Curse of Dimensionality: Number of Samples  Suppose we want to use the nearest neighbor approach with k = 1 (1NN)  This feature is not discriminative, i.e. it does not  separate the classes well  Suppose we start with only one feature  We decide to use 2 features. For the 1NN method to work well, need a lot of samples, i.e. samples have to be dense  To maintain the same density as in 1D (9 samples per unit length), how many samples do we need?
  • 8. Curse of Dimensionality: Number of Samples  We need 92 samples to maintain the same density as in 1D
  • 9. Curse of Dimensionality: Number of Samples  Of course, when we go from 1 feature to 2, no one gives us more samples, we still have 9  This is way too sparse for 1NN to work well
  • 10. Curse of Dimensionality: Number of Samples  Things go from bad to worse if we decide to use 3 features:  If 9 was dense enough in 1D, in 3D we need 93=729 samples!
  • 11. Curse of Dimensionality: Number of Samples  In general, if n samples is dense enough in 1D  Then in d dimensions we need n d samples!  And n d grows really really fast as a function of d  Common pitfall:  If we can’t solve a problem with a few features, adding more features seems like a good idea  However the number of samples usually stays the same  The method with more features is likely to perform worse instead of expected better
  • 12. Curse of Dimensionality: Number of Samples  For a fixed number of samples, as we add features, the graph of classification error:  Thus for each fixed sample size n, there is the optimal number of features to use
  • 13. The Curse of Dimensionality  We should try to avoid creating lot of features  Often no choice, problem starts with many features  Example: Face Detection  One sample point is k by m array of pixels  Feature extraction is not trivial, usually every  pixel is taken as a feature  Typical dimension is 20 by 20 = 400  Suppose 10 samples are dense enough for 1 dimension. Need only 10400samples
  • 14. The Curse of Dimensionality  Face Detection, dimension of one sample point is km  The fact that we set up the problem with km dimensions (features) does not mean it is really a km-dimensional problem  Most likely we are not setting the problem up with the right features  If we used better features, we are likely need much less than km-dimensions  Space of all k by m images has km dimensions  Space of all k by m faces must be much smaller, since faces form a tiny fraction of all possible images
  • 15. Dimensionality Reduction  We want to reduce the number of dimensions because:  Efficiency.  Measurement costs.  Storage costs.  Computation costs.  Classification performance.  Ease of interpretation/modeling.
  • 16. Principal Components  The idea is to project onto the subspace which accounts for most of the variance.  This is accomplished by projecting onto the eigenvectors of the covariance matrix associated with the largest eigenvalues.  This is generally not the projection best suited for classification.  It can be a useful method for providing a first-cut reduction in dimension from a high dimensional space
  • 17. Feature Combination as a method to reduce Dimensionality  High dimensionality is challenging and redundant  It is natural to try to reduce dimensionality  Reduce dimensionality by feature combination: combine old features x to create new features y  For Example,  Ideally, the new vector y should retain from x all information important for classification
  • 18. Principle Component Analysis (PCA)  Main idea: seek most accurate data representation in a lower dimensional space  Example in 2-D ◦ Project data to 1-D subspace (a line) which minimize the projection error  Notice that the the good line to use for projection lies in the direction of largest variance
  • 19. PCA: Approximation of Elliptical Cloud in 3D
  • 20. Thank You! The End.

Notes de l'éditeur

  1. Let&apos;s say you have a straight line 100 yards long and you dropped a penny somewhere on it. It wouldn&apos;t be too hard to find. You walk along the line and it takes two minutes.Now let&apos;s say you have a square 100 yards across and you dropped a penny somewhere on it. It would be pretty hard, like searching across two football fields stuck together. It could take days.Now a cube 100 yards across. That&apos;s like searching a 30-story building the size of a football stadium. Ugh.The difficulty of searching through the space gets a *lot* harder as you have more dimensions. You might not realize this intuitively when it&apos;s just stated in mathematical formulas, since they all have the same &quot;width&quot;. That&apos;s the curse of dimensionality. It gets to have a name because it is unintuitive, useful, and yet simple.
  2. Let the complete probability space for one variable be represented by the unit interval (0, 1), and imagine drawing ten samples along that interval. Each sample would then have to represent 10% of the probability space (on average). Now consider a second variable defined on another, orthogonal, (0, 1) interval, also being represented by ten samples. We have produced 10 (x1 , x2 ) points on a plane defined by the orthogonal x1 , x2 lines, and representing a new probability space. But the new space has 10 ? 10 = 100 area units, so each of the ten points now represents only 1% of the probability space. It would require 100 points for each point to represent the same 10% of the probability space that was represented by 10 points in only one dimension. This is illustrated below, where the 10 points are not drawn at random to illustrate their diminished coverage.The Number of Points Would Need to Increase Exponentially to Maintain a Given Accuracy.Now, although there are 100 (x1 , x2 ) points, neither axis has more accuracy than was provided by the previous 10, because there are 10 values of x1 required for the 10 different values of x2 necessary to represent the new probability plane. Because in practice the samples are taken at random it appears as though there are more samples, but this isn&apos;t the case in terms of probability space coverage which is related to the density of points, not the number of observations per axis.Next consider adding another dimension creating a probability space represented by a cube with ten units on a side, and 1000 probability units within. It now requires stacking 10 of the previous (x1 , x2 ) planes to construct the new (x1 , x2 , x3 ) cube, and of course 10 times the number of samples per axis to maintain the former level of accuracy, or probability coverage. It is obvious that10n samples would be required for a n-dimension problem.