Soumettre la recherche
Mettre en ligne
Cutting edge of Machine Learning
•
Télécharger en tant que PPTX, PDF
•
3 j'aime
•
992 vues
IT Weekend
Suivre
by Sergii Shelpuk
Lire moins
Lire la suite
Technologie
Formation
Signaler
Partager
Signaler
Partager
1 sur 23
Télécharger maintenant
Recommandé
Deep Learning A-Z™: Artificial Neural Networks (ANN) - Module 1
Deep Learning A-Z™: Artificial Neural Networks (ANN) - Module 1
Kirill Eremenko
20130912_mit _geordie_rose
20130912_mit _geordie_rose
Geordie Rose
20130911 idc hpc_geordie_rose_final
20130911 idc hpc_geordie_rose_final
Geordie Rose
Deep Learning for Developers (October 2017)
Deep Learning for Developers (October 2017)
Julien SIMON
Auto-Encoders and PCA, a brief psychological background
Auto-Encoders and PCA, a brief psychological background
Amgad Muhammad
Simple Introduction to AutoEncoder
Simple Introduction to AutoEncoder
Jun Lang
More Data, More Problems: Evolving big data machine learning pipelines with S...
More Data, More Problems: Evolving big data machine learning pipelines with S...
Alex Sadovsky
Managing data workflows with Luigi
Managing data workflows with Luigi
Teemu Kurppa
Recommandé
Deep Learning A-Z™: Artificial Neural Networks (ANN) - Module 1
Deep Learning A-Z™: Artificial Neural Networks (ANN) - Module 1
Kirill Eremenko
20130912_mit _geordie_rose
20130912_mit _geordie_rose
Geordie Rose
20130911 idc hpc_geordie_rose_final
20130911 idc hpc_geordie_rose_final
Geordie Rose
Deep Learning for Developers (October 2017)
Deep Learning for Developers (October 2017)
Julien SIMON
Auto-Encoders and PCA, a brief psychological background
Auto-Encoders and PCA, a brief psychological background
Amgad Muhammad
Simple Introduction to AutoEncoder
Simple Introduction to AutoEncoder
Jun Lang
More Data, More Problems: Evolving big data machine learning pipelines with S...
More Data, More Problems: Evolving big data machine learning pipelines with S...
Alex Sadovsky
Managing data workflows with Luigi
Managing data workflows with Luigi
Teemu Kurppa
Deep Learning for Computer Vision: Data Augmentation (UPC 2016)
Deep Learning for Computer Vision: Data Augmentation (UPC 2016)
Universitat Politècnica de Catalunya
Unsupervised Feature Learning
Unsupervised Feature Learning
Amgad Muhammad
Introduction to deep learning in python and Matlab
Introduction to deep learning in python and Matlab
Imry Kissos
Intro to Deep learning - Autoencoders
Intro to Deep learning - Autoencoders
Akash Goel
進化するWebトラッキングの話 #ssmjp
進化するWebトラッキングの話 #ssmjp
sonickun
Autoencoders for image_classification
Autoencoders for image_classification
Cenk Bircanoğlu
Electricity price forecasting with Recurrent Neural Networks
Electricity price forecasting with Recurrent Neural Networks
Taegyun Jeon
Robust and declarative machine learning pipelines for predictive buying at Ba...
Robust and declarative machine learning pipelines for predictive buying at Ba...
Gianmario Spacagna
Unsupervised Computer Vision: The Current State of the Art
Unsupervised Computer Vision: The Current State of the Art
TJ Torres
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
Simplilearn
Deep Learning for Computer Vision (1/4): Image Analytics @ laSalle 2016
Deep Learning for Computer Vision (1/4): Image Analytics @ laSalle 2016
Universitat Politècnica de Catalunya
Neural network in matlab
Neural network in matlab
Fahim Khan
20200723_insight_release_plan
20200723_insight_release_plan
Jamie (Taka) Wang
Deep Learning Hardware: Past, Present, & Future
Deep Learning Hardware: Past, Present, & Future
Rouyun Pan
Quality attributes testing. From Architecture to test acceptance
Quality attributes testing. From Architecture to test acceptance
IT Weekend
Mobile development for JavaScript developer
Mobile development for JavaScript developer
IT Weekend
Building an Innovation & Strategy Process
Building an Innovation & Strategy Process
IT Weekend
IT Professionals – The Right Time/The Right Place
IT Professionals – The Right Time/The Right Place
IT Weekend
Building a Data Driven Organization
Building a Data Driven Organization
IT Weekend
7 Tools for the Product Owner
7 Tools for the Product Owner
IT Weekend
Hacking your Doorbell
Hacking your Doorbell
IT Weekend
An era of possibilities, a window in time
An era of possibilities, a window in time
IT Weekend
Contenu connexe
En vedette
Deep Learning for Computer Vision: Data Augmentation (UPC 2016)
Deep Learning for Computer Vision: Data Augmentation (UPC 2016)
Universitat Politècnica de Catalunya
Unsupervised Feature Learning
Unsupervised Feature Learning
Amgad Muhammad
Introduction to deep learning in python and Matlab
Introduction to deep learning in python and Matlab
Imry Kissos
Intro to Deep learning - Autoencoders
Intro to Deep learning - Autoencoders
Akash Goel
進化するWebトラッキングの話 #ssmjp
進化するWebトラッキングの話 #ssmjp
sonickun
Autoencoders for image_classification
Autoencoders for image_classification
Cenk Bircanoğlu
Electricity price forecasting with Recurrent Neural Networks
Electricity price forecasting with Recurrent Neural Networks
Taegyun Jeon
Robust and declarative machine learning pipelines for predictive buying at Ba...
Robust and declarative machine learning pipelines for predictive buying at Ba...
Gianmario Spacagna
En vedette
(8)
Deep Learning for Computer Vision: Data Augmentation (UPC 2016)
Deep Learning for Computer Vision: Data Augmentation (UPC 2016)
Unsupervised Feature Learning
Unsupervised Feature Learning
Introduction to deep learning in python and Matlab
Introduction to deep learning in python and Matlab
Intro to Deep learning - Autoencoders
Intro to Deep learning - Autoencoders
進化するWebトラッキングの話 #ssmjp
進化するWebトラッキングの話 #ssmjp
Autoencoders for image_classification
Autoencoders for image_classification
Electricity price forecasting with Recurrent Neural Networks
Electricity price forecasting with Recurrent Neural Networks
Robust and declarative machine learning pipelines for predictive buying at Ba...
Robust and declarative machine learning pipelines for predictive buying at Ba...
Similaire à Cutting edge of Machine Learning
Unsupervised Computer Vision: The Current State of the Art
Unsupervised Computer Vision: The Current State of the Art
TJ Torres
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
Simplilearn
Deep Learning for Computer Vision (1/4): Image Analytics @ laSalle 2016
Deep Learning for Computer Vision (1/4): Image Analytics @ laSalle 2016
Universitat Politècnica de Catalunya
Neural network in matlab
Neural network in matlab
Fahim Khan
20200723_insight_release_plan
20200723_insight_release_plan
Jamie (Taka) Wang
Deep Learning Hardware: Past, Present, & Future
Deep Learning Hardware: Past, Present, & Future
Rouyun Pan
Similaire à Cutting edge of Machine Learning
(6)
Unsupervised Computer Vision: The Current State of the Art
Unsupervised Computer Vision: The Current State of the Art
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
Deep Learning for Computer Vision (1/4): Image Analytics @ laSalle 2016
Deep Learning for Computer Vision (1/4): Image Analytics @ laSalle 2016
Neural network in matlab
Neural network in matlab
20200723_insight_release_plan
20200723_insight_release_plan
Deep Learning Hardware: Past, Present, & Future
Deep Learning Hardware: Past, Present, & Future
Plus de IT Weekend
Quality attributes testing. From Architecture to test acceptance
Quality attributes testing. From Architecture to test acceptance
IT Weekend
Mobile development for JavaScript developer
Mobile development for JavaScript developer
IT Weekend
Building an Innovation & Strategy Process
Building an Innovation & Strategy Process
IT Weekend
IT Professionals – The Right Time/The Right Place
IT Professionals – The Right Time/The Right Place
IT Weekend
Building a Data Driven Organization
Building a Data Driven Organization
IT Weekend
7 Tools for the Product Owner
7 Tools for the Product Owner
IT Weekend
Hacking your Doorbell
Hacking your Doorbell
IT Weekend
An era of possibilities, a window in time
An era of possibilities, a window in time
IT Weekend
Web services automation from sketch
Web services automation from sketch
IT Weekend
Why Ruby?
Why Ruby?
IT Weekend
REST that won't make you cry
REST that won't make you cry
IT Weekend
Как договариваться с начальником и заказчиком: выбираем нужный протокол общения
Как договариваться с начальником и заказчиком: выбираем нужный протокол общения
IT Weekend
Обзор программы SAP HANA Startup Focus
Обзор программы SAP HANA Startup Focus
IT Weekend
World of Agile: Kanban
World of Agile: Kanban
IT Weekend
Risk Management
Risk Management
IT Weekend
«Spring Integration as Integration Patterns Provider»
«Spring Integration as Integration Patterns Provider»
IT Weekend
Parallel Programming In Modern World .NET Technics
Parallel Programming In Modern World .NET Technics
IT Weekend
Parallel programming in modern world .net technics shared
Parallel programming in modern world .net technics shared
IT Weekend
Maximize Effectiveness of Human Capital
Maximize Effectiveness of Human Capital
IT Weekend
“Using C#/.NET – “Controversial Topics & Common Mistakes”
“Using C#/.NET – “Controversial Topics & Common Mistakes”
IT Weekend
Plus de IT Weekend
(20)
Quality attributes testing. From Architecture to test acceptance
Quality attributes testing. From Architecture to test acceptance
Mobile development for JavaScript developer
Mobile development for JavaScript developer
Building an Innovation & Strategy Process
Building an Innovation & Strategy Process
IT Professionals – The Right Time/The Right Place
IT Professionals – The Right Time/The Right Place
Building a Data Driven Organization
Building a Data Driven Organization
7 Tools for the Product Owner
7 Tools for the Product Owner
Hacking your Doorbell
Hacking your Doorbell
An era of possibilities, a window in time
An era of possibilities, a window in time
Web services automation from sketch
Web services automation from sketch
Why Ruby?
Why Ruby?
REST that won't make you cry
REST that won't make you cry
Как договариваться с начальником и заказчиком: выбираем нужный протокол общения
Как договариваться с начальником и заказчиком: выбираем нужный протокол общения
Обзор программы SAP HANA Startup Focus
Обзор программы SAP HANA Startup Focus
World of Agile: Kanban
World of Agile: Kanban
Risk Management
Risk Management
«Spring Integration as Integration Patterns Provider»
«Spring Integration as Integration Patterns Provider»
Parallel Programming In Modern World .NET Technics
Parallel Programming In Modern World .NET Technics
Parallel programming in modern world .net technics shared
Parallel programming in modern world .net technics shared
Maximize Effectiveness of Human Capital
Maximize Effectiveness of Human Capital
“Using C#/.NET – “Controversial Topics & Common Mistakes”
“Using C#/.NET – “Controversial Topics & Common Mistakes”
Dernier
"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan
Fwdays
Story boards and shot lists for my a level piece
Story boards and shot lists for my a level piece
charlottematthew16
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck Presentation
Slibray Presentation
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
Fwdays
Unleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding Club
Kalema Edgar
Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!
Manik S Magar
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Safe Software
DevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platforms
Sergiu Bodiu
Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!
Commit University
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial Buildings
Memoori
"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko
Fwdays
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
null - The Open Security Community
WordPress Websites for Engineers: Elevate Your Brand
WordPress Websites for Engineers: Elevate Your Brand
gvaughan
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
Fwdays
Developer Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQL
ScyllaDB
DMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special Edition
Dubai Multi Commodity Centre
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdf
Addepto
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdf
Alex Barbosa Coqueiro
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
Wonjun Hwang
What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024
Stephanie Beckett
Dernier
(20)
"ML in Production",Oleksandr Bagan
"ML in Production",Oleksandr Bagan
Story boards and shot lists for my a level piece
Story boards and shot lists for my a level piece
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck Presentation
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
Unleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding Club
Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
DevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platforms
Nell’iperspazio con Rocket: il Framework Web di Rust!
Nell’iperspazio con Rocket: il Framework Web di Rust!
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial Buildings
"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
WordPress Websites for Engineers: Elevate Your Brand
WordPress Websites for Engineers: Elevate Your Brand
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
Developer Data Modeling Mistakes: From Postgres to NoSQL
Developer Data Modeling Mistakes: From Postgres to NoSQL
DMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special Edition
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdf
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdf
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
Bun (KitWorks Team Study 노별마루 발표 2024.4.22)
What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024
Cutting edge of Machine Learning
1.
Machine Learning The Cutting
Edge Sergii Shelpuk Director, Data Science SoftServe, Inc. sshel@softserveinc.com
2.
Classification Problem Recognize what
is a bike and what is a moon
3.
Classification Problem Classifier ©A. Ng
4.
Classification Problem pixel intensity
5.
Classification Problem Raw data
does not represent the picture well. You need some smart features contains wheels contains seas
6.
Feature Extraction Classifier Featureextractor ©A. Ng
7.
Feature Extraction Can we
do better?
8.
Neural Networks a a a a a a a a a a a a a a features bike moon
9.
Neural Networks
10.
Neural Networks aX a0 a1 a2 w0 w1 w2 Activation function: aX
= f(a0, a1, a2, w0, w1, w2) Example (logistic): aX = 1 / (1 + e-(a0*w0+a1*w1+a2*w2))
11.
Autoencoder
12.
Autoencoder © H. Lee
et al.
13.
Autoencoder © Q Le
et al.
14.
Deep Learning Neural
Network Pre-trained as Autoencoder Typical classification neural network Moon
15.
Deep Learning Neural
NetworkVideoText/NLPImages ©A. Ng
16.
Deep Learning Neural
Network Hints and Tips Using unlabeled data Avoiding overfitting Computational efficiency
17.
Using Unlabeled Data wheels handlebar
18.
Avoiding Overfitting Sparsity constraint
limits variance of autoencoder
19.
Avoiding Overfitting Dropout ensures
generalization of the neural network
20.
Computational Efficiency Thousands
of cores Base Clock: 300-900 MHz Memory: 2-6 Gb Performance: up to 3.5 Tflops Instruction-level parallelism Shared memory Up to 4 devices in cluster GPU computing provides cheapest computational power
21.
Feature Learning: MNIST Data: Features:
22.
Feature Learning: Galaxy
Zoo Data: Features:
23.
Thank you!
Télécharger maintenant