SlideShare uma empresa Scribd logo
1 de 21
Financial time series forecasting
using support
vector machines
Author: Kyoung-jae Kim
2003 Elsevier B.V.
Outline
•
•
•
•

Introduction to SVM
Introduction to datasets
Experimental settings
Analysis of experimental results
Linear separability
• Linear separability
– In general, two groups are linearly separable in ndimensional space if they can be separated by an
(n − 1)-dimensional hyperplane.
Support Vector Machines
• Maximum-margin hyperplane
maximum-margin hyperplane
Formalization
• Training data
• Hyperplane
• Parallel bounding hyperplanes
Objective
• Minimize (in w, b)
||w||

• subject to (for any i=1, …, n)
A 2-D case
• In 2-D:
– Training data:
xi

ci

<1, 1> 1
<2, 2> 1
<2, 1> -1
<3, 2> -1

-2x+2y+1=1

-2x+2y+1=0
-2x+2y+1=-1

w=<-2, 2>
b=-1
margin=sqrt(2)/2
Not linear separable
• No hyperplane can separate the two groups
Soft Margin
• Choose a hyperplane that splits the examples
as cleanly as possible
• Still maximizing the distance to the nearest
cleanly split examples
• Introduce an error cost C
d*C
Higher dimensions
• Separation might be easier
Kernel Trick
• Build maximal margin hyperplanes in highdimenisonal feature space depends on inner
product: more cost
• Use a kernel function that lives in low
dimensions, but behaves like an inner product
in high dimensions
Kernels
• Polynomial
– K(p, q) = (p•q + c)d

• Radial basis function
– K(p, q) = exp(-γ||p-q||2)

• Gaussian radial basis
– K(p, q) = exp(-||p-q||2/2δ2)
Tuning parameters
• Error weight
–C

• Kernel parameters
– δ2
–d
– c0
Underfitting & Overfitting
• Underfitting
• Overfitting
• High generalization ability
Datasets
• Input variables
– 12 technical indicators

• Target attribute
– Korea composite stock price index (KOSPI)

• 2928 trading days
– 80% for training, 20% for holdout
Settings (1/3)
• SVM
– kernels
• polynomial kernel
• Gaussian radial basis function
– δ2

– error cost C
Settings (2/3)
• BP-Network
– layers
• 3

– number of hidden nodes
• 6, 12, 24

– learning epochs per training example
• 50, 100, 200

– learning rate
• 0.1

– momentum
• 0.1

– input nodes
• 12
Settings (3/3)
• Case-Based Reasoning
– k-NN
• k = 1, 2, 3, 4, 5

– distance evaluation
• Euclidean distance
Experimental results
• The results of SVMs with various C where δ2 is fixed
at 25
• Too small C
• underfitting*

• Too large C
• overfitting*

* F.E.H. Tay, L. Cao, Application of support vector machines in -nancial time series forecasting, Omega 29 (2001) 309–317
Experimental results
• The results of SVMs with various δ2 where C is fixed
at 78
• Small value of δ2
• overfitting*

• Large value of δ2
• underfitting*

* F.E.H. Tay, L. Cao, Application of support vector machines in -nancial time series forecasting, Omega 29 (2001) 309–317
Experimental results and conclusion
• SVM outperformes BPN and CBR
• SVM minimizes structural risk
• SVM provides a promising alternative for
financial time-series forecasting
• Issues
– parameter tuning

Mais conteúdo relacionado

Mais procurados

Orthogonal Frequency Division Multiplexing (OFDM)
Orthogonal Frequency Division Multiplexing (OFDM)Orthogonal Frequency Division Multiplexing (OFDM)
Orthogonal Frequency Division Multiplexing (OFDM)ashishsoni1505
 
A brief history of cloud computing
A brief history of cloud computingA brief history of cloud computing
A brief history of cloud computingOneserve
 
Evolution of the cloud
Evolution of the cloudEvolution of the cloud
Evolution of the cloudsagaroceanic11
 
Packet radio protocol
Packet radio protocolPacket radio protocol
Packet radio protocolPriya Kaushal
 
Deep Learning Based Voice Activity Detection and Speech Enhancement
Deep Learning Based Voice Activity Detection and Speech EnhancementDeep Learning Based Voice Activity Detection and Speech Enhancement
Deep Learning Based Voice Activity Detection and Speech EnhancementNAVER Engineering
 
2.5G, second and half generation, All about 2.5..
2.5G, second and half generation, All about 2.5..2.5G, second and half generation, All about 2.5..
2.5G, second and half generation, All about 2.5..Muhammad Ahad
 
Wavelength division multiplexing
Wavelength division multiplexingWavelength division multiplexing
Wavelength division multiplexingNalin Dubey
 
Report on telecom sector
Report on telecom sectorReport on telecom sector
Report on telecom sectorJonty Mohta
 
History and Evolution of Cloud computing (Safaricom cloud)
History and Evolution of Cloud computing (Safaricom cloud)History and Evolution of Cloud computing (Safaricom cloud)
History and Evolution of Cloud computing (Safaricom cloud)Ben Wakhungu
 
Internet Traffic Forecasting using Time Series Methods
Internet Traffic Forecasting using Time Series MethodsInternet Traffic Forecasting using Time Series Methods
Internet Traffic Forecasting using Time Series MethodsAjay Ohri
 
Addressing PNT threats in critical defense infrastructure
Addressing PNT threats in critical defense infrastructureAddressing PNT threats in critical defense infrastructure
Addressing PNT threats in critical defense infrastructureADVA
 
Ericsson hds 8000 wp 16
Ericsson hds 8000 wp 16Ericsson hds 8000 wp 16
Ericsson hds 8000 wp 16Mainstay
 

Mais procurados (20)

Orthogonal Frequency Division Multiplexing (OFDM)
Orthogonal Frequency Division Multiplexing (OFDM)Orthogonal Frequency Division Multiplexing (OFDM)
Orthogonal Frequency Division Multiplexing (OFDM)
 
A brief history of cloud computing
A brief history of cloud computingA brief history of cloud computing
A brief history of cloud computing
 
Evolution of the cloud
Evolution of the cloudEvolution of the cloud
Evolution of the cloud
 
Packet radio protocol
Packet radio protocolPacket radio protocol
Packet radio protocol
 
Deep Learning Based Voice Activity Detection and Speech Enhancement
Deep Learning Based Voice Activity Detection and Speech EnhancementDeep Learning Based Voice Activity Detection and Speech Enhancement
Deep Learning Based Voice Activity Detection and Speech Enhancement
 
2.5G, second and half generation, All about 2.5..
2.5G, second and half generation, All about 2.5..2.5G, second and half generation, All about 2.5..
2.5G, second and half generation, All about 2.5..
 
CloudStack Architecture
CloudStack ArchitectureCloudStack Architecture
CloudStack Architecture
 
Cloud Computing: Virtualization
Cloud Computing: VirtualizationCloud Computing: Virtualization
Cloud Computing: Virtualization
 
Wavelength division multiplexing
Wavelength division multiplexingWavelength division multiplexing
Wavelength division multiplexing
 
Report on telecom sector
Report on telecom sectorReport on telecom sector
Report on telecom sector
 
scheduling techniques and SLA.pptx
scheduling techniques and SLA.pptxscheduling techniques and SLA.pptx
scheduling techniques and SLA.pptx
 
History and Evolution of Cloud computing (Safaricom cloud)
History and Evolution of Cloud computing (Safaricom cloud)History and Evolution of Cloud computing (Safaricom cloud)
History and Evolution of Cloud computing (Safaricom cloud)
 
Broadband isdn and atm
Broadband  isdn and atmBroadband  isdn and atm
Broadband isdn and atm
 
WiMAX vs LTE
WiMAX vs LTEWiMAX vs LTE
WiMAX vs LTE
 
Internet Traffic Forecasting using Time Series Methods
Internet Traffic Forecasting using Time Series MethodsInternet Traffic Forecasting using Time Series Methods
Internet Traffic Forecasting using Time Series Methods
 
Mobile computing unit-I-notes 07.01.2020
Mobile computing unit-I-notes 07.01.2020Mobile computing unit-I-notes 07.01.2020
Mobile computing unit-I-notes 07.01.2020
 
Grid computing
Grid computingGrid computing
Grid computing
 
Addressing PNT threats in critical defense infrastructure
Addressing PNT threats in critical defense infrastructureAddressing PNT threats in critical defense infrastructure
Addressing PNT threats in critical defense infrastructure
 
Ericsson hds 8000 wp 16
Ericsson hds 8000 wp 16Ericsson hds 8000 wp 16
Ericsson hds 8000 wp 16
 
Ofdma Basics
Ofdma BasicsOfdma Basics
Ofdma Basics
 

Semelhante a Time series Forecasting using svm

Compressed learning for time series classification
Compressed learning for time series classificationCompressed learning for time series classification
Compressed learning for time series classification學翰 施
 
“Fundamentals of Training AI Models for Computer Vision Applications,” a Pres...
“Fundamentals of Training AI Models for Computer Vision Applications,” a Pres...“Fundamentals of Training AI Models for Computer Vision Applications,” a Pres...
“Fundamentals of Training AI Models for Computer Vision Applications,” a Pres...Edge AI and Vision Alliance
 
Presentation on SOM
Presentation on SOMPresentation on SOM
Presentation on SOMArchiLab 7
 
Efficient anomaly detection via matrix sketching
Efficient anomaly detection via matrix sketchingEfficient anomaly detection via matrix sketching
Efficient anomaly detection via matrix sketchingHsing-chuan Hsieh
 
Machine Learning workshop by GDSC Amity University Chhattisgarh
Machine Learning workshop by GDSC Amity University ChhattisgarhMachine Learning workshop by GDSC Amity University Chhattisgarh
Machine Learning workshop by GDSC Amity University ChhattisgarhPoorabpatel
 
Classification using Apache SystemML by Prithviraj Sen
Classification using Apache SystemML by Prithviraj SenClassification using Apache SystemML by Prithviraj Sen
Classification using Apache SystemML by Prithviraj SenArvind Surve
 
Classification using Apache SystemML by Prithviraj Sen
Classification using Apache SystemML by Prithviraj SenClassification using Apache SystemML by Prithviraj Sen
Classification using Apache SystemML by Prithviraj SenArvind Surve
 
Svm map reduce_slides
Svm map reduce_slidesSvm map reduce_slides
Svm map reduce_slidesSara Asher
 
Afsar ml applied_svm
Afsar ml applied_svmAfsar ml applied_svm
Afsar ml applied_svmUmmeHaniAsif
 
Hardware Acceleration for Machine Learning
Hardware Acceleration for Machine LearningHardware Acceleration for Machine Learning
Hardware Acceleration for Machine LearningCastLabKAIST
 
background.pptx
background.pptxbackground.pptx
background.pptxKabileshCm
 
svm-proyekt.pptx
svm-proyekt.pptxsvm-proyekt.pptx
svm-proyekt.pptxElinEliyev
 
Huong dan cu the svm
Huong dan cu the svmHuong dan cu the svm
Huong dan cu the svmtaikhoan262
 

Semelhante a Time series Forecasting using svm (20)

Compressed learning for time series classification
Compressed learning for time series classificationCompressed learning for time series classification
Compressed learning for time series classification
 
Support Vector Machines
Support Vector MachinesSupport Vector Machines
Support Vector Machines
 
“Fundamentals of Training AI Models for Computer Vision Applications,” a Pres...
“Fundamentals of Training AI Models for Computer Vision Applications,” a Pres...“Fundamentals of Training AI Models for Computer Vision Applications,” a Pres...
“Fundamentals of Training AI Models for Computer Vision Applications,” a Pres...
 
Presentation on SOM
Presentation on SOMPresentation on SOM
Presentation on SOM
 
Efficient anomaly detection via matrix sketching
Efficient anomaly detection via matrix sketchingEfficient anomaly detection via matrix sketching
Efficient anomaly detection via matrix sketching
 
Support Vector Machines ( SVM )
Support Vector Machines ( SVM ) Support Vector Machines ( SVM )
Support Vector Machines ( SVM )
 
Machine Learning workshop by GDSC Amity University Chhattisgarh
Machine Learning workshop by GDSC Amity University ChhattisgarhMachine Learning workshop by GDSC Amity University Chhattisgarh
Machine Learning workshop by GDSC Amity University Chhattisgarh
 
Classification using Apache SystemML by Prithviraj Sen
Classification using Apache SystemML by Prithviraj SenClassification using Apache SystemML by Prithviraj Sen
Classification using Apache SystemML by Prithviraj Sen
 
Classification using Apache SystemML by Prithviraj Sen
Classification using Apache SystemML by Prithviraj SenClassification using Apache SystemML by Prithviraj Sen
Classification using Apache SystemML by Prithviraj Sen
 
Svm map reduce_slides
Svm map reduce_slidesSvm map reduce_slides
Svm map reduce_slides
 
BIRTE-13-Kawashima
BIRTE-13-KawashimaBIRTE-13-Kawashima
BIRTE-13-Kawashima
 
Afsar ml applied_svm
Afsar ml applied_svmAfsar ml applied_svm
Afsar ml applied_svm
 
Hardware Acceleration for Machine Learning
Hardware Acceleration for Machine LearningHardware Acceleration for Machine Learning
Hardware Acceleration for Machine Learning
 
background.pptx
background.pptxbackground.pptx
background.pptx
 
svm-proyekt.pptx
svm-proyekt.pptxsvm-proyekt.pptx
svm-proyekt.pptx
 
Guide
GuideGuide
Guide
 
Huong dan cu the svm
Huong dan cu the svmHuong dan cu the svm
Huong dan cu the svm
 
.doc
.doc.doc
.doc
 
.doc
.doc.doc
.doc
 
.doc
.doc.doc
.doc
 

Mais de Institute of Technology Telkom

Mais de Institute of Technology Telkom (20)

Econopysics
EconopysicsEconopysics
Econopysics
 
Science and religion 100622120615-phpapp01
Science and religion 100622120615-phpapp01Science and religion 100622120615-phpapp01
Science and religion 100622120615-phpapp01
 
Konvergensi sains dan_spiritualitas
Konvergensi sains dan_spiritualitasKonvergensi sains dan_spiritualitas
Konvergensi sains dan_spiritualitas
 
Matematika arah kiblat mikrajuddin abdullah 2017
Matematika arah kiblat   mikrajuddin abdullah 2017Matematika arah kiblat   mikrajuddin abdullah 2017
Matematika arah kiblat mikrajuddin abdullah 2017
 
Iau solar effects 2005
Iau solar effects 2005Iau solar effects 2005
Iau solar effects 2005
 
Hfmsilri2jun14
Hfmsilri2jun14Hfmsilri2jun14
Hfmsilri2jun14
 
Fisika komputasi
Fisika komputasiFisika komputasi
Fisika komputasi
 
Computer Aided Process Planning
Computer Aided Process PlanningComputer Aided Process Planning
Computer Aided Process Planning
 
Archimedes
ArchimedesArchimedes
Archimedes
 
Web and text
Web and textWeb and text
Web and text
 
Web data mining
Web data miningWeb data mining
Web data mining
 
Timeseries forecasting
Timeseries forecastingTimeseries forecasting
Timeseries forecasting
 
Fuzzy logic
Fuzzy logicFuzzy logic
Fuzzy logic
 
World population 1950--2050
World population 1950--2050World population 1950--2050
World population 1950--2050
 
neural networks
 neural networks neural networks
neural networks
 
Artificial neural networks
Artificial neural networks Artificial neural networks
Artificial neural networks
 
002 ray modeling dynamic systems
002 ray modeling dynamic systems002 ray modeling dynamic systems
002 ray modeling dynamic systems
 
002 ray modeling dynamic systems
002 ray modeling dynamic systems002 ray modeling dynamic systems
002 ray modeling dynamic systems
 
System dynamics majors fair
System dynamics majors fairSystem dynamics majors fair
System dynamics majors fair
 
System dynamics math representation
System dynamics math representationSystem dynamics math representation
System dynamics math representation
 

Último

mini mental status format.docx
mini    mental       status     format.docxmini    mental       status     format.docx
mini mental status format.docxPoojaSen20
 
Student login on Anyboli platform.helpin
Student login on Anyboli platform.helpinStudent login on Anyboli platform.helpin
Student login on Anyboli platform.helpinRaunakKeshri1
 
A Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformA Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformChameera Dedduwage
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfsanyamsingh5019
 
1029-Danh muc Sach Giao Khoa khoi 6.pdf
1029-Danh muc Sach Giao Khoa khoi  6.pdf1029-Danh muc Sach Giao Khoa khoi  6.pdf
1029-Danh muc Sach Giao Khoa khoi 6.pdfQucHHunhnh
 
How to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxHow to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxmanuelaromero2013
 
Activity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfActivity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfciinovamais
 
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Sapana Sha
 
Separation of Lanthanides/ Lanthanides and Actinides
Separation of Lanthanides/ Lanthanides and ActinidesSeparation of Lanthanides/ Lanthanides and Actinides
Separation of Lanthanides/ Lanthanides and ActinidesFatimaKhan178732
 
CARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxCARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxGaneshChakor2
 
Accessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactAccessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactdawncurless
 
Z Score,T Score, Percential Rank and Box Plot Graph
Z Score,T Score, Percential Rank and Box Plot GraphZ Score,T Score, Percential Rank and Box Plot Graph
Z Score,T Score, Percential Rank and Box Plot GraphThiyagu K
 
Introduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxIntroduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxpboyjonauth
 
Industrial Policy - 1948, 1956, 1973, 1977, 1980, 1991
Industrial Policy - 1948, 1956, 1973, 1977, 1980, 1991Industrial Policy - 1948, 1956, 1973, 1977, 1980, 1991
Industrial Policy - 1948, 1956, 1973, 1977, 1980, 1991RKavithamani
 
Employee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxEmployee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxNirmalaLoungPoorunde1
 
Measures of Central Tendency: Mean, Median and Mode
Measures of Central Tendency: Mean, Median and ModeMeasures of Central Tendency: Mean, Median and Mode
Measures of Central Tendency: Mean, Median and ModeThiyagu K
 

Último (20)

mini mental status format.docx
mini    mental       status     format.docxmini    mental       status     format.docx
mini mental status format.docx
 
Student login on Anyboli platform.helpin
Student login on Anyboli platform.helpinStudent login on Anyboli platform.helpin
Student login on Anyboli platform.helpin
 
TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
TataKelola dan KamSiber Kecerdasan Buatan v022.pdfTataKelola dan KamSiber Kecerdasan Buatan v022.pdf
TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
 
A Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformA Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy Reform
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdf
 
1029-Danh muc Sach Giao Khoa khoi 6.pdf
1029-Danh muc Sach Giao Khoa khoi  6.pdf1029-Danh muc Sach Giao Khoa khoi  6.pdf
1029-Danh muc Sach Giao Khoa khoi 6.pdf
 
Mattingly "AI & Prompt Design: Structured Data, Assistants, & RAG"
Mattingly "AI & Prompt Design: Structured Data, Assistants, & RAG"Mattingly "AI & Prompt Design: Structured Data, Assistants, & RAG"
Mattingly "AI & Prompt Design: Structured Data, Assistants, & RAG"
 
Código Creativo y Arte de Software | Unidad 1
Código Creativo y Arte de Software | Unidad 1Código Creativo y Arte de Software | Unidad 1
Código Creativo y Arte de Software | Unidad 1
 
How to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxHow to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptx
 
Activity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfActivity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdf
 
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptxINDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
 
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
 
Separation of Lanthanides/ Lanthanides and Actinides
Separation of Lanthanides/ Lanthanides and ActinidesSeparation of Lanthanides/ Lanthanides and Actinides
Separation of Lanthanides/ Lanthanides and Actinides
 
CARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxCARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptx
 
Accessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactAccessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impact
 
Z Score,T Score, Percential Rank and Box Plot Graph
Z Score,T Score, Percential Rank and Box Plot GraphZ Score,T Score, Percential Rank and Box Plot Graph
Z Score,T Score, Percential Rank and Box Plot Graph
 
Introduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptxIntroduction to AI in Higher Education_draft.pptx
Introduction to AI in Higher Education_draft.pptx
 
Industrial Policy - 1948, 1956, 1973, 1977, 1980, 1991
Industrial Policy - 1948, 1956, 1973, 1977, 1980, 1991Industrial Policy - 1948, 1956, 1973, 1977, 1980, 1991
Industrial Policy - 1948, 1956, 1973, 1977, 1980, 1991
 
Employee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxEmployee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptx
 
Measures of Central Tendency: Mean, Median and Mode
Measures of Central Tendency: Mean, Median and ModeMeasures of Central Tendency: Mean, Median and Mode
Measures of Central Tendency: Mean, Median and Mode
 

Time series Forecasting using svm

  • 1. Financial time series forecasting using support vector machines Author: Kyoung-jae Kim 2003 Elsevier B.V.
  • 2. Outline • • • • Introduction to SVM Introduction to datasets Experimental settings Analysis of experimental results
  • 3. Linear separability • Linear separability – In general, two groups are linearly separable in ndimensional space if they can be separated by an (n − 1)-dimensional hyperplane.
  • 4. Support Vector Machines • Maximum-margin hyperplane maximum-margin hyperplane
  • 5. Formalization • Training data • Hyperplane • Parallel bounding hyperplanes
  • 6. Objective • Minimize (in w, b) ||w|| • subject to (for any i=1, …, n)
  • 7. A 2-D case • In 2-D: – Training data: xi ci <1, 1> 1 <2, 2> 1 <2, 1> -1 <3, 2> -1 -2x+2y+1=1 -2x+2y+1=0 -2x+2y+1=-1 w=<-2, 2> b=-1 margin=sqrt(2)/2
  • 8. Not linear separable • No hyperplane can separate the two groups
  • 9. Soft Margin • Choose a hyperplane that splits the examples as cleanly as possible • Still maximizing the distance to the nearest cleanly split examples • Introduce an error cost C d*C
  • 11. Kernel Trick • Build maximal margin hyperplanes in highdimenisonal feature space depends on inner product: more cost • Use a kernel function that lives in low dimensions, but behaves like an inner product in high dimensions
  • 12. Kernels • Polynomial – K(p, q) = (p•q + c)d • Radial basis function – K(p, q) = exp(-γ||p-q||2) • Gaussian radial basis – K(p, q) = exp(-||p-q||2/2δ2)
  • 13. Tuning parameters • Error weight –C • Kernel parameters – δ2 –d – c0
  • 14. Underfitting & Overfitting • Underfitting • Overfitting • High generalization ability
  • 15. Datasets • Input variables – 12 technical indicators • Target attribute – Korea composite stock price index (KOSPI) • 2928 trading days – 80% for training, 20% for holdout
  • 16. Settings (1/3) • SVM – kernels • polynomial kernel • Gaussian radial basis function – δ2 – error cost C
  • 17. Settings (2/3) • BP-Network – layers • 3 – number of hidden nodes • 6, 12, 24 – learning epochs per training example • 50, 100, 200 – learning rate • 0.1 – momentum • 0.1 – input nodes • 12
  • 18. Settings (3/3) • Case-Based Reasoning – k-NN • k = 1, 2, 3, 4, 5 – distance evaluation • Euclidean distance
  • 19. Experimental results • The results of SVMs with various C where δ2 is fixed at 25 • Too small C • underfitting* • Too large C • overfitting* * F.E.H. Tay, L. Cao, Application of support vector machines in -nancial time series forecasting, Omega 29 (2001) 309–317
  • 20. Experimental results • The results of SVMs with various δ2 where C is fixed at 78 • Small value of δ2 • overfitting* • Large value of δ2 • underfitting* * F.E.H. Tay, L. Cao, Application of support vector machines in -nancial time series forecasting, Omega 29 (2001) 309–317
  • 21. Experimental results and conclusion • SVM outperformes BPN and CBR • SVM minimizes structural risk • SVM provides a promising alternative for financial time-series forecasting • Issues – parameter tuning