Deep learning: the future of recommendations

Balázs Hidasi
Balázs HidasiHead of Data Mining and Research at Gravity R&D
Deep learning:
the future of recommendations
Balázs Hidasi
Head of Data Mining and Research
Gravity meetup @ Startup Safary
April 21, 2016
Deep learning in the headlines
Deep learning in the background
• Life improving services
 Speech recognition
 Personal assistants (e.g. Siri,
Cortana)
 Computer vision, object
recognition
 Machine translation
 Chatbot technology
 Natural Language Processing
 Face recognition
 Self driving cars
• For fun
 Text generation
 Composing music
 Painting pictures
 Etc.
What is deep learning?
• A class of machine learning algorithms
 that use a cascade of multiple non-linear processing layers
 and complex model structures
 to learn different representations of the data in each layer
 where higher level features are derived from lower level
features
 to form a hierarchical representation.
Deep learning is not a new topic
• First deep network proposed in the 1970s
• More papers in the 80s and 90s
• Why now?
 Older research was not used widely in practice
 Applications were much more simplistic that today’s
Neural networks: a brief overview
Neurons, neural networks
• Neuron: rough abstraction of the human neuron
 Receives inputs (signals)
 Sum weighted inputs is big enough  signal
 Amplifiers and inhibitors
 Basic pattern recognition
• Neural network: neurons connected to one another
• Feedforward networks: neurons are organized into
layers
 Connections only between subsequent layers
𝑦
𝑥1
𝑥2
𝑥3
𝑥4
𝑓(. )
𝑖=1
𝑁
𝑤𝑖 𝑥𝑖 + 𝑏
𝑥1
𝑥2
𝑥3
ℎ1
1
ℎ2
1
ℎ3
1
ℎ1
2
ℎ2
2
Networks that big enough: go deep not wide
• Feedforward neural networks are universal
approximators
 Can imitate any function if they are big enough
 (Also needs enough in-out pairs to learn)
• What is big enough?
 Number of layers / neurons
 Theoretical „big enough” conditions massively overshoot
• Go deep, not wide
 The number of neurons required for good approximation is
polynomial in the input if the network is deep enough
 Otherwise it is exponential
Training neural networks
• Forward pass: get the current estimate of the target
o 𝑠𝑗
1
= 𝑖 𝑤𝑖,𝑗
1
𝑥𝑖 + 𝑏𝑗
1
; ℎ𝑗
1
= 𝑓 𝑠𝑗
1
o 𝑠 𝑘
2
= 𝑗 𝑤𝑗,𝑘
2
ℎ𝑗
1
+ 𝑏 𝑘
2
; ℎ 𝑘
2
= 𝑓 𝑠 𝑘
2
o …
o 𝑠𝑙
𝑂
= 𝑘 𝑤 𝑘,𝑙
𝑁+1
ℎ 𝑘
𝑁
+ 𝑏𝑙
𝑂
; 𝑦𝑙 = 𝑓 𝑠𝑙
𝑂
• Backward pass: correct weights to reduce error
 Gradient descentLayer Error Gradient
(w.r.t. weights between current and prev. layer)
Output Defined loss
(e.g. 𝐿 = 𝑖=1
𝑁 𝑜
𝑦𝑖 − 𝑦𝑖
2
)
𝜕𝐿
𝜕𝑤𝑗,𝑖
(𝑁+1)
=
𝜕𝐿
𝜕𝑦𝑖
∗
𝜕𝑦𝑖
𝜕𝑠𝑖
𝑂 ∗
𝜕𝑠𝑖
𝑂
𝜕𝑤𝑗,𝑖
𝑁+1 =
𝜕𝐿
𝜕𝑦𝑖
𝑓′
𝑠𝑖
𝑂
ℎ𝑗
𝑁
𝑁 𝑡ℎ
hidden
𝛿𝑖
𝑁
=
𝜕𝐿
𝜕𝑦𝑖
∗
𝜕𝑦𝑖
𝜕𝑠𝑖
𝑂
𝜕𝐿
𝜕𝑤 𝑘,𝑗
𝑁 =
𝑖
𝜕𝐿
𝜕𝑦𝑖
∗
𝜕𝑦𝑖
𝜕𝑠𝑖
𝑂 ∗
𝜕𝑠𝑖
𝑂
𝜕ℎ𝑗
𝑁 ∗
𝜕ℎ𝑗
𝑁
𝜕𝑠𝑗
𝑁 ∗
𝜕𝑠𝑗
𝑁
𝜕𝑤𝑗,𝑖
𝑁 =
𝑖
𝛿𝑖
𝑁
𝑤𝑖,𝑗
𝑁+1
𝑓′ 𝑠𝑗
𝑁
ℎ 𝑘
𝑁−1
(𝑁 −
𝛿𝑗
𝑁−1
=
𝑖
𝛿𝑖
𝑁
𝑤𝑖,𝑗
𝑁+1
𝑓′ 𝑠𝑗
𝑁
𝜕𝐿
𝜕𝑤𝑙,𝑘
𝑁−1 =
𝑗
𝛿𝑗
𝑁−1
𝑤𝑗,𝑘
𝑁
𝑓′ 𝑠 𝑘
𝑁−1
ℎ𝑙
𝑁−2
…
1 𝑠𝑡
hidden
𝛿 𝑘
1 𝜕𝐿
𝜕𝑤𝑖,𝑗
1 =
𝑘
𝛿 𝑘
1
𝑤 𝑘,𝑙
2
𝑓′
𝑠𝑗
1
𝑥𝑖
Challenges of training deep networks
• Saturation
• Vanishing gradients
• Overfitting
• Slowness of second order methods
• Slow convergence, stucks in local optima with first
order methods
• (Exploding gradients)
Why now?
Breakthroughs in research
• Saturation & vanishing gradients
 Layer-by-layer training (2006)
 Non-saturating activation functions, e.g. ReLU (2013)
• Overfitting
 Dropout (2014)
• Convergence problems
 Adagrad, Adadelta, Adam, RMSProp, etc.
Computational power
• Natural increase in computational power
• GP GPU technology
Intermission
Don’t give in to the HYPE
• Deep learning is impressive but
 deep learning is not true AI
o it may be a component of it when
and if AI is created
 deep learning is not how the human
brain works
 95% of machine learning tasks don’t
require deep learning
 deep learning requires a lot of
computational power
• Deep learning is a tool
 which is successful in certain,
previously very challenging domains
(speech recognition, computer
vision, NLP, etc.)
 that excels in pattern recognition
You are here
Deep learning for RecSys
From the Netflix prize...
• Netflix prize (2006-2009)
 Gave a huge push to recommender systems research
 Determined the direction of research for years
 Task:
o Some (User, Item, Rating) known triplets
o (User, Item) pairs with unknown rating
o Predict the missing ratings (1-5)
... to recommenders in practice
• Ratings  events [implicit feedback]
 Lots of services don’t allow for rating
 Majority of users don’t rate
 Monitored passively  preferences have to be infered
• Rating prediction  ranking [top N recommendations]
 All that matters is the relevancy of the top N items
 Rating prediction is biased
• User  session / situation [session-based / context-driven
recommendation]
 Users are not logged in, identification is unreliable
 Accounts used by multiple users
 Aim of the session (e.g. buy a good laptop)
 Similar behavior of different users in a situation, different behavior of the same
user in different situations
Challenges in RecSys
• Session modeling
 Most of the algorithms are personalized
 A few are item-to-item
o Recommends similar items
o Also used for session-based recommendations (industry de facto standard)
 There are no good session based solutions
• Incorporating factors that influence user clicks
 Users click based on what they see
o Title
o Product image
o Description
 and on their knowledge of the product
o Usually harder to model
o Except when the product is content (e.g. music)
Deep learning to the rescue – Session modeling
• Recurrent Neural Networks (RNN)
 Sequence modeling
 Hidden state: next state is based on the previous hidden state and the current input
 „Infinite” depth
 More sophisticated versions: GRU, LSTM
• Needs to be adapted to the recommendation task
• GRU4Rec:
 Session-parallel minibatch training for handling the large variance in session lengths
 Sampling the output for reasonable training times, without losing much accuracy
 Ranking loss for better item ranking
• Results: 15-30% improvement over item-to-item recommendations
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
RSC15 VIDEO
Recall@20
Item-kNN
GRU4Rec
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
RSC15 VIDEO
MRR@20
Item-kNN
GRU4Rec
Other uses of deep learning for recsys
• Incorporating content directly
 Music, images, video, text
 User influencing aspects of the items
 Direct content representation
• Context-state modeling from sensory data
 IoT devices
 Lot of sensory data
 Some missing and noise
 Infer context state and recommend accordingly
• Interactive recommenders using chatbots
• Personalized content generation
 Today’s news
 Images in personalized style with personalized content
• Etc...
There is work to be done
• DL + RecSys research: just started
 Last year:
o 0 long papers, 1 short paper and 1 poster that is loosely connected
 This year:
o 10+ submissions to RecSys in this topic
o DLRS 2016 workshop @ RecSys
• Open questions
 (More) Application areas
 Adaptations required for the recsys problem
 Scalability
 Best practices
 ...
Thanks for your attention!
1 of 23

Recommended

Deep Learning for Recommender Systems RecSys2017 Tutorial by
Deep Learning for Recommender Systems RecSys2017 Tutorial Deep Learning for Recommender Systems RecSys2017 Tutorial
Deep Learning for Recommender Systems RecSys2017 Tutorial Alexandros Karatzoglou
32.2K views80 slides
User Behavior Analytics Using Machine Learning by
User Behavior Analytics Using Machine LearningUser Behavior Analytics Using Machine Learning
User Behavior Analytics Using Machine LearningDNIF
423 views20 slides
Introduction to Recurrent Neural Network by
Introduction to Recurrent Neural NetworkIntroduction to Recurrent Neural Network
Introduction to Recurrent Neural NetworkYan Xu
4.8K views32 slides
Deep Learning in Recommender Systems - RecSys Summer School 2017 by
Deep Learning in Recommender Systems - RecSys Summer School 2017Deep Learning in Recommender Systems - RecSys Summer School 2017
Deep Learning in Recommender Systems - RecSys Summer School 2017Balázs Hidasi
11.3K views77 slides
Deep Learning for Recommender Systems by
Deep Learning for Recommender SystemsDeep Learning for Recommender Systems
Deep Learning for Recommender SystemsJustin Basilico
21K views35 slides
[기초개념] Recurrent Neural Network (RNN) 소개 by
[기초개념] Recurrent Neural Network (RNN) 소개[기초개념] Recurrent Neural Network (RNN) 소개
[기초개념] Recurrent Neural Network (RNN) 소개Donghyeon Kim
5.9K views62 slides

More Related Content

What's hot

Recent Trends in Personalization at Netflix by
Recent Trends in Personalization at NetflixRecent Trends in Personalization at Netflix
Recent Trends in Personalization at NetflixJustin Basilico
24.2K views57 slides
Introduction For seq2seq(sequence to sequence) and RNN by
Introduction For seq2seq(sequence to sequence) and RNNIntroduction For seq2seq(sequence to sequence) and RNN
Introduction For seq2seq(sequence to sequence) and RNNHye-min Ahn
3.9K views20 slides
Recommendation systems by
Recommendation systemsRecommendation systems
Recommendation systemsSaurabhWani6
260 views15 slides
Wasserstein GAN 수학 이해하기 I by
Wasserstein GAN 수학 이해하기 IWasserstein GAN 수학 이해하기 I
Wasserstein GAN 수학 이해하기 ISungbin Lim
37.7K views97 slides
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys... by
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...Xavier Amatriain
16.5K views46 slides
Recsys 2014 Tutorial - The Recommender Problem Revisited by
Recsys 2014 Tutorial - The Recommender Problem RevisitedRecsys 2014 Tutorial - The Recommender Problem Revisited
Recsys 2014 Tutorial - The Recommender Problem RevisitedXavier Amatriain
27K views112 slides

What's hot(20)

Recent Trends in Personalization at Netflix by Justin Basilico
Recent Trends in Personalization at NetflixRecent Trends in Personalization at Netflix
Recent Trends in Personalization at Netflix
Justin Basilico24.2K views
Introduction For seq2seq(sequence to sequence) and RNN by Hye-min Ahn
Introduction For seq2seq(sequence to sequence) and RNNIntroduction For seq2seq(sequence to sequence) and RNN
Introduction For seq2seq(sequence to sequence) and RNN
Hye-min Ahn3.9K views
Recommendation systems by SaurabhWani6
Recommendation systemsRecommendation systems
Recommendation systems
SaurabhWani6260 views
Wasserstein GAN 수학 이해하기 I by Sungbin Lim
Wasserstein GAN 수학 이해하기 IWasserstein GAN 수학 이해하기 I
Wasserstein GAN 수학 이해하기 I
Sungbin Lim37.7K views
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys... by Xavier Amatriain
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Recsys 2016 tutorial: Lessons learned from building real-life recommender sys...
Xavier Amatriain16.5K views
Recsys 2014 Tutorial - The Recommender Problem Revisited by Xavier Amatriain
Recsys 2014 Tutorial - The Recommender Problem RevisitedRecsys 2014 Tutorial - The Recommender Problem Revisited
Recsys 2014 Tutorial - The Recommender Problem Revisited
Xavier Amatriain27K views
Machine Learning for Q&A Sites: The Quora Example by Xavier Amatriain
Machine Learning for Q&A Sites: The Quora ExampleMachine Learning for Q&A Sites: The Quora Example
Machine Learning for Q&A Sites: The Quora Example
Xavier Amatriain6.2K views
1시간만에 GAN(Generative Adversarial Network) 완전 정복하기 by NAVER Engineering
1시간만에 GAN(Generative Adversarial Network) 완전 정복하기1시간만에 GAN(Generative Adversarial Network) 완전 정복하기
1시간만에 GAN(Generative Adversarial Network) 완전 정복하기
NAVER Engineering23.1K views
Recommendation system by Ding Li
Recommendation systemRecommendation system
Recommendation system
Ding Li315 views
Deeper Things: How Netflix Leverages Deep Learning in Recommendations and Se... by Sudeep Das, Ph.D.
 Deeper Things: How Netflix Leverages Deep Learning in Recommendations and Se... Deeper Things: How Netflix Leverages Deep Learning in Recommendations and Se...
Deeper Things: How Netflix Leverages Deep Learning in Recommendations and Se...
Sudeep Das, Ph.D.13K views
Natural language processing and transformer models by Ding Li
Natural language processing and transformer modelsNatural language processing and transformer models
Natural language processing and transformer models
Ding Li600 views
Tutorial on Deep Learning in Recommender System, Lars summer school 2019 by Anoop Deoras
Tutorial on Deep Learning in Recommender System, Lars summer school 2019Tutorial on Deep Learning in Recommender System, Lars summer school 2019
Tutorial on Deep Learning in Recommender System, Lars summer school 2019
Anoop Deoras2.2K views
Recent advances in deep recommender systems by NAVER Engineering
Recent advances in deep recommender systemsRecent advances in deep recommender systems
Recent advances in deep recommender systems
NAVER Engineering1.9K views
The fundamentals of Machine Learning by hichem felouat
The fundamentals of Machine LearningThe fundamentals of Machine Learning
The fundamentals of Machine Learning
hichem felouat968 views
Learning to Rank for Recommender Systems - ACM RecSys 2013 tutorial by Alexandros Karatzoglou
Learning to Rank for Recommender Systems -  ACM RecSys 2013 tutorialLearning to Rank for Recommender Systems -  ACM RecSys 2013 tutorial
Learning to Rank for Recommender Systems - ACM RecSys 2013 tutorial
PPT4: Frameworks & Libraries of Machine Learning & Deep Learning by akira-ai
PPT4: Frameworks & Libraries of Machine Learning & Deep Learning PPT4: Frameworks & Libraries of Machine Learning & Deep Learning
PPT4: Frameworks & Libraries of Machine Learning & Deep Learning
akira-ai380 views

Similar to Deep learning: the future of recommendations

Deep learning to the rescue - solving long standing problems of recommender ... by
Deep learning to the rescue - solving long standing problems of recommender ...Deep learning to the rescue - solving long standing problems of recommender ...
Deep learning to the rescue - solving long standing problems of recommender ...Balázs Hidasi
13.6K views17 slides
MDEC Data Matters Series: machine learning and Deep Learning, A Primer by
MDEC Data Matters Series: machine learning and Deep Learning, A PrimerMDEC Data Matters Series: machine learning and Deep Learning, A Primer
MDEC Data Matters Series: machine learning and Deep Learning, A PrimerPoo Kuan Hoong
1.1K views55 slides
Deep learning - a primer by
Deep learning - a primerDeep learning - a primer
Deep learning - a primerUwe Friedrichsen
2.8K views137 slides
Deep learning - a primer by
Deep learning - a primerDeep learning - a primer
Deep learning - a primerShirin Elsinghorst
4.7K views137 slides
Introduction to deep learning by
Introduction to deep learningIntroduction to deep learning
Introduction to deep learningAmr Rashed
362 views68 slides
Big Data Malaysia - A Primer on Deep Learning by
Big Data Malaysia - A Primer on Deep LearningBig Data Malaysia - A Primer on Deep Learning
Big Data Malaysia - A Primer on Deep LearningPoo Kuan Hoong
696 views53 slides

Similar to Deep learning: the future of recommendations(20)

Deep learning to the rescue - solving long standing problems of recommender ... by Balázs Hidasi
Deep learning to the rescue - solving long standing problems of recommender ...Deep learning to the rescue - solving long standing problems of recommender ...
Deep learning to the rescue - solving long standing problems of recommender ...
Balázs Hidasi13.6K views
MDEC Data Matters Series: machine learning and Deep Learning, A Primer by Poo Kuan Hoong
MDEC Data Matters Series: machine learning and Deep Learning, A PrimerMDEC Data Matters Series: machine learning and Deep Learning, A Primer
MDEC Data Matters Series: machine learning and Deep Learning, A Primer
Poo Kuan Hoong1.1K views
Introduction to deep learning by Amr Rashed
Introduction to deep learningIntroduction to deep learning
Introduction to deep learning
Amr Rashed362 views
Big Data Malaysia - A Primer on Deep Learning by Poo Kuan Hoong
Big Data Malaysia - A Primer on Deep LearningBig Data Malaysia - A Primer on Deep Learning
Big Data Malaysia - A Primer on Deep Learning
Poo Kuan Hoong696 views
An Introduction to Deep Learning by Poo Kuan Hoong
An Introduction to Deep LearningAn Introduction to Deep Learning
An Introduction to Deep Learning
Poo Kuan Hoong11.8K views
Deep learning tutorial 9/2019 by Amr Rashed
Deep learning tutorial 9/2019Deep learning tutorial 9/2019
Deep learning tutorial 9/2019
Amr Rashed587 views
Deep Learning Tutorial by Amr Rashed
Deep Learning TutorialDeep Learning Tutorial
Deep Learning Tutorial
Amr Rashed5.4K views
Deep learning introduction by Adwait Bhave
Deep learning introductionDeep learning introduction
Deep learning introduction
Adwait Bhave382 views
DEF CON 24 - Clarence Chio - machine duping 101 by Felipe Prado
DEF CON 24 - Clarence Chio - machine duping 101DEF CON 24 - Clarence Chio - machine duping 101
DEF CON 24 - Clarence Chio - machine duping 101
Felipe Prado70 views
Deep Learning, an interactive introduction for NLP-ers by Roelof Pieters
Deep Learning, an interactive introduction for NLP-ersDeep Learning, an interactive introduction for NLP-ers
Deep Learning, an interactive introduction for NLP-ers
Roelof Pieters13.7K views
Using Deep Learning to do Real-Time Scoring in Practical Applications by Greg Makowski
Using Deep Learning to do Real-Time Scoring in Practical ApplicationsUsing Deep Learning to do Real-Time Scoring in Practical Applications
Using Deep Learning to do Real-Time Scoring in Practical Applications
Greg Makowski4.5K views
Visualization of Deep Learning by YaminiAlapati1
Visualization of Deep LearningVisualization of Deep Learning
Visualization of Deep Learning
YaminiAlapati1158 views
Introduction to deep learning by doppenhe
Introduction to deep learningIntroduction to deep learning
Introduction to deep learning
doppenhe782 views
Deep Learning: a birds eye view by Roelof Pieters
Deep Learning: a birds eye viewDeep Learning: a birds eye view
Deep Learning: a birds eye view
Roelof Pieters8.3K views
Separating Hype from Reality in Deep Learning with Sameer Farooqui by Databricks
 Separating Hype from Reality in Deep Learning with Sameer Farooqui Separating Hype from Reality in Deep Learning with Sameer Farooqui
Separating Hype from Reality in Deep Learning with Sameer Farooqui
Databricks970 views

More from Balázs Hidasi

Egyedi termék kreatívok tömeges gyártása generatív AI segítségével by
Egyedi termék kreatívok tömeges gyártása generatív AI segítségévelEgyedi termék kreatívok tömeges gyártása generatív AI segítségével
Egyedi termék kreatívok tömeges gyártása generatív AI segítségévelBalázs Hidasi
45 views57 slides
The Effect of Third Party Implementations on Reproducibility by
The Effect of Third Party Implementations on ReproducibilityThe Effect of Third Party Implementations on Reproducibility
The Effect of Third Party Implementations on ReproducibilityBalázs Hidasi
113 views41 slides
GRU4Rec v2 - Recurrent Neural Networks with Top-k Gains for Session-based Rec... by
GRU4Rec v2 - Recurrent Neural Networks with Top-k Gains for Session-based Rec...GRU4Rec v2 - Recurrent Neural Networks with Top-k Gains for Session-based Rec...
GRU4Rec v2 - Recurrent Neural Networks with Top-k Gains for Session-based Rec...Balázs Hidasi
3.2K views30 slides
Parallel Recurrent Neural Network Architectures for Feature-rich Session-base... by
Parallel Recurrent Neural Network Architectures for Feature-rich Session-base...Parallel Recurrent Neural Network Architectures for Feature-rich Session-base...
Parallel Recurrent Neural Network Architectures for Feature-rich Session-base...Balázs Hidasi
2.3K views15 slides
Context aware factorization methods for implicit feedback based recommendatio... by
Context aware factorization methods for implicit feedback based recommendatio...Context aware factorization methods for implicit feedback based recommendatio...
Context aware factorization methods for implicit feedback based recommendatio...Balázs Hidasi
524 views34 slides
Context-aware preference modeling with factorization by
Context-aware preference modeling with factorizationContext-aware preference modeling with factorization
Context-aware preference modeling with factorizationBalázs Hidasi
1.7K views22 slides

More from Balázs Hidasi(15)

Egyedi termék kreatívok tömeges gyártása generatív AI segítségével by Balázs Hidasi
Egyedi termék kreatívok tömeges gyártása generatív AI segítségévelEgyedi termék kreatívok tömeges gyártása generatív AI segítségével
Egyedi termék kreatívok tömeges gyártása generatív AI segítségével
Balázs Hidasi45 views
The Effect of Third Party Implementations on Reproducibility by Balázs Hidasi
The Effect of Third Party Implementations on ReproducibilityThe Effect of Third Party Implementations on Reproducibility
The Effect of Third Party Implementations on Reproducibility
Balázs Hidasi113 views
GRU4Rec v2 - Recurrent Neural Networks with Top-k Gains for Session-based Rec... by Balázs Hidasi
GRU4Rec v2 - Recurrent Neural Networks with Top-k Gains for Session-based Rec...GRU4Rec v2 - Recurrent Neural Networks with Top-k Gains for Session-based Rec...
GRU4Rec v2 - Recurrent Neural Networks with Top-k Gains for Session-based Rec...
Balázs Hidasi3.2K views
Parallel Recurrent Neural Network Architectures for Feature-rich Session-base... by Balázs Hidasi
Parallel Recurrent Neural Network Architectures for Feature-rich Session-base...Parallel Recurrent Neural Network Architectures for Feature-rich Session-base...
Parallel Recurrent Neural Network Architectures for Feature-rich Session-base...
Balázs Hidasi2.3K views
Context aware factorization methods for implicit feedback based recommendatio... by Balázs Hidasi
Context aware factorization methods for implicit feedback based recommendatio...Context aware factorization methods for implicit feedback based recommendatio...
Context aware factorization methods for implicit feedback based recommendatio...
Balázs Hidasi524 views
Context-aware preference modeling with factorization by Balázs Hidasi
Context-aware preference modeling with factorizationContext-aware preference modeling with factorization
Context-aware preference modeling with factorization
Balázs Hidasi1.7K views
Approximate modeling of continuous context in factorization algorithms (CaRR1... by Balázs Hidasi
Approximate modeling of continuous context in factorization algorithms (CaRR1...Approximate modeling of continuous context in factorization algorithms (CaRR1...
Approximate modeling of continuous context in factorization algorithms (CaRR1...
Balázs Hidasi1.4K views
Utilizing additional information in factorization methods (research overview,... by Balázs Hidasi
Utilizing additional information in factorization methods (research overview,...Utilizing additional information in factorization methods (research overview,...
Utilizing additional information in factorization methods (research overview,...
Balázs Hidasi628 views
Az implicit ajánlási probléma és néhány megoldása (BME TMIT szeminárium előad... by Balázs Hidasi
Az implicit ajánlási probléma és néhány megoldása (BME TMIT szeminárium előad...Az implicit ajánlási probléma és néhány megoldása (BME TMIT szeminárium előad...
Az implicit ajánlási probléma és néhány megoldása (BME TMIT szeminárium előad...
Balázs Hidasi383 views
Context-aware similarities within the factorization framework (CaRR 2013 pres... by Balázs Hidasi
Context-aware similarities within the factorization framework (CaRR 2013 pres...Context-aware similarities within the factorization framework (CaRR 2013 pres...
Context-aware similarities within the factorization framework (CaRR 2013 pres...
Balázs Hidasi1.2K views
iTALS: implicit tensor factorization for context-aware recommendations (ECML/... by Balázs Hidasi
iTALS: implicit tensor factorization for context-aware recommendations (ECML/...iTALS: implicit tensor factorization for context-aware recommendations (ECML/...
iTALS: implicit tensor factorization for context-aware recommendations (ECML/...
Balázs Hidasi637 views
Initialization of matrix factorization (CaRR 2012 presentation) by Balázs Hidasi
Initialization of matrix factorization (CaRR 2012 presentation)Initialization of matrix factorization (CaRR 2012 presentation)
Initialization of matrix factorization (CaRR 2012 presentation)
Balázs Hidasi737 views
ShiftTree: model alapú idősor-osztályozó (VK 2009 előadás) by Balázs Hidasi
ShiftTree: model alapú idősor-osztályozó (VK 2009 előadás)ShiftTree: model alapú idősor-osztályozó (VK 2009 előadás)
ShiftTree: model alapú idősor-osztályozó (VK 2009 előadás)
Balázs Hidasi327 views
ShiftTree: model alapú idősor-osztályozó (ML@BP előadás, 2012) by Balázs Hidasi
ShiftTree: model alapú idősor-osztályozó (ML@BP előadás, 2012)ShiftTree: model alapú idősor-osztályozó (ML@BP előadás, 2012)
ShiftTree: model alapú idősor-osztályozó (ML@BP előadás, 2012)
Balázs Hidasi357 views
ShiftTree: model based time series classifier (ECML/PKDD 2011 presentation) by Balázs Hidasi
ShiftTree: model based time series classifier (ECML/PKDD 2011 presentation)ShiftTree: model based time series classifier (ECML/PKDD 2011 presentation)
ShiftTree: model based time series classifier (ECML/PKDD 2011 presentation)
Balázs Hidasi725 views

Recently uploaded

HTTP headers that make your website go faster - devs.gent November 2023 by
HTTP headers that make your website go faster - devs.gent November 2023HTTP headers that make your website go faster - devs.gent November 2023
HTTP headers that make your website go faster - devs.gent November 2023Thijs Feryn
19 views151 slides
AMAZON PRODUCT RESEARCH.pdf by
AMAZON PRODUCT RESEARCH.pdfAMAZON PRODUCT RESEARCH.pdf
AMAZON PRODUCT RESEARCH.pdfJerikkLaureta
15 views13 slides
ChatGPT and AI for Web Developers by
ChatGPT and AI for Web DevelopersChatGPT and AI for Web Developers
ChatGPT and AI for Web DevelopersMaximiliano Firtman
181 views82 slides
Voice Logger - Telephony Integration Solution at Aegis by
Voice Logger - Telephony Integration Solution at AegisVoice Logger - Telephony Integration Solution at Aegis
Voice Logger - Telephony Integration Solution at AegisNirmal Sharma
17 views1 slide
Attacking IoT Devices from a Web Perspective - Linux Day by
Attacking IoT Devices from a Web Perspective - Linux Day Attacking IoT Devices from a Web Perspective - Linux Day
Attacking IoT Devices from a Web Perspective - Linux Day Simone Onofri
15 views68 slides
1st parposal presentation.pptx by
1st parposal presentation.pptx1st parposal presentation.pptx
1st parposal presentation.pptxi238212
9 views3 slides

Recently uploaded(20)

HTTP headers that make your website go faster - devs.gent November 2023 by Thijs Feryn
HTTP headers that make your website go faster - devs.gent November 2023HTTP headers that make your website go faster - devs.gent November 2023
HTTP headers that make your website go faster - devs.gent November 2023
Thijs Feryn19 views
AMAZON PRODUCT RESEARCH.pdf by JerikkLaureta
AMAZON PRODUCT RESEARCH.pdfAMAZON PRODUCT RESEARCH.pdf
AMAZON PRODUCT RESEARCH.pdf
JerikkLaureta15 views
Voice Logger - Telephony Integration Solution at Aegis by Nirmal Sharma
Voice Logger - Telephony Integration Solution at AegisVoice Logger - Telephony Integration Solution at Aegis
Voice Logger - Telephony Integration Solution at Aegis
Nirmal Sharma17 views
Attacking IoT Devices from a Web Perspective - Linux Day by Simone Onofri
Attacking IoT Devices from a Web Perspective - Linux Day Attacking IoT Devices from a Web Perspective - Linux Day
Attacking IoT Devices from a Web Perspective - Linux Day
Simone Onofri15 views
1st parposal presentation.pptx by i238212
1st parposal presentation.pptx1st parposal presentation.pptx
1st parposal presentation.pptx
i2382129 views
6g - REPORT.pdf by Liveplex
6g - REPORT.pdf6g - REPORT.pdf
6g - REPORT.pdf
Liveplex9 views
TouchLog: Finger Micro Gesture Recognition Using Photo-Reflective Sensors by sugiuralab
TouchLog: Finger Micro Gesture Recognition  Using Photo-Reflective SensorsTouchLog: Finger Micro Gesture Recognition  Using Photo-Reflective Sensors
TouchLog: Finger Micro Gesture Recognition Using Photo-Reflective Sensors
sugiuralab15 views
Igniting Next Level Productivity with AI-Infused Data Integration Workflows by Safe Software
Igniting Next Level Productivity with AI-Infused Data Integration Workflows Igniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration Workflows
Safe Software225 views
Piloting & Scaling Successfully With Microsoft Viva by Richard Harbridge
Piloting & Scaling Successfully With Microsoft VivaPiloting & Scaling Successfully With Microsoft Viva
Piloting & Scaling Successfully With Microsoft Viva
The details of description: Techniques, tips, and tangents on alternative tex... by BookNet Canada
The details of description: Techniques, tips, and tangents on alternative tex...The details of description: Techniques, tips, and tangents on alternative tex...
The details of description: Techniques, tips, and tangents on alternative tex...
BookNet Canada121 views
Perth MeetUp November 2023 by Michael Price
Perth MeetUp November 2023 Perth MeetUp November 2023
Perth MeetUp November 2023
Michael Price15 views
DALI Basics Course 2023 by Ivory Egg
DALI Basics Course  2023DALI Basics Course  2023
DALI Basics Course 2023
Ivory Egg14 views
handbook for web 3 adoption.pdf by Liveplex
handbook for web 3 adoption.pdfhandbook for web 3 adoption.pdf
handbook for web 3 adoption.pdf
Liveplex19 views

Deep learning: the future of recommendations

  • 1. Deep learning: the future of recommendations Balázs Hidasi Head of Data Mining and Research Gravity meetup @ Startup Safary April 21, 2016
  • 2. Deep learning in the headlines
  • 3. Deep learning in the background • Life improving services  Speech recognition  Personal assistants (e.g. Siri, Cortana)  Computer vision, object recognition  Machine translation  Chatbot technology  Natural Language Processing  Face recognition  Self driving cars • For fun  Text generation  Composing music  Painting pictures  Etc.
  • 4. What is deep learning? • A class of machine learning algorithms  that use a cascade of multiple non-linear processing layers  and complex model structures  to learn different representations of the data in each layer  where higher level features are derived from lower level features  to form a hierarchical representation.
  • 5. Deep learning is not a new topic • First deep network proposed in the 1970s • More papers in the 80s and 90s • Why now?  Older research was not used widely in practice  Applications were much more simplistic that today’s
  • 6. Neural networks: a brief overview
  • 7. Neurons, neural networks • Neuron: rough abstraction of the human neuron  Receives inputs (signals)  Sum weighted inputs is big enough  signal  Amplifiers and inhibitors  Basic pattern recognition • Neural network: neurons connected to one another • Feedforward networks: neurons are organized into layers  Connections only between subsequent layers 𝑦 𝑥1 𝑥2 𝑥3 𝑥4 𝑓(. ) 𝑖=1 𝑁 𝑤𝑖 𝑥𝑖 + 𝑏 𝑥1 𝑥2 𝑥3 ℎ1 1 ℎ2 1 ℎ3 1 ℎ1 2 ℎ2 2
  • 8. Networks that big enough: go deep not wide • Feedforward neural networks are universal approximators  Can imitate any function if they are big enough  (Also needs enough in-out pairs to learn) • What is big enough?  Number of layers / neurons  Theoretical „big enough” conditions massively overshoot • Go deep, not wide  The number of neurons required for good approximation is polynomial in the input if the network is deep enough  Otherwise it is exponential
  • 9. Training neural networks • Forward pass: get the current estimate of the target o 𝑠𝑗 1 = 𝑖 𝑤𝑖,𝑗 1 𝑥𝑖 + 𝑏𝑗 1 ; ℎ𝑗 1 = 𝑓 𝑠𝑗 1 o 𝑠 𝑘 2 = 𝑗 𝑤𝑗,𝑘 2 ℎ𝑗 1 + 𝑏 𝑘 2 ; ℎ 𝑘 2 = 𝑓 𝑠 𝑘 2 o … o 𝑠𝑙 𝑂 = 𝑘 𝑤 𝑘,𝑙 𝑁+1 ℎ 𝑘 𝑁 + 𝑏𝑙 𝑂 ; 𝑦𝑙 = 𝑓 𝑠𝑙 𝑂 • Backward pass: correct weights to reduce error  Gradient descentLayer Error Gradient (w.r.t. weights between current and prev. layer) Output Defined loss (e.g. 𝐿 = 𝑖=1 𝑁 𝑜 𝑦𝑖 − 𝑦𝑖 2 ) 𝜕𝐿 𝜕𝑤𝑗,𝑖 (𝑁+1) = 𝜕𝐿 𝜕𝑦𝑖 ∗ 𝜕𝑦𝑖 𝜕𝑠𝑖 𝑂 ∗ 𝜕𝑠𝑖 𝑂 𝜕𝑤𝑗,𝑖 𝑁+1 = 𝜕𝐿 𝜕𝑦𝑖 𝑓′ 𝑠𝑖 𝑂 ℎ𝑗 𝑁 𝑁 𝑡ℎ hidden 𝛿𝑖 𝑁 = 𝜕𝐿 𝜕𝑦𝑖 ∗ 𝜕𝑦𝑖 𝜕𝑠𝑖 𝑂 𝜕𝐿 𝜕𝑤 𝑘,𝑗 𝑁 = 𝑖 𝜕𝐿 𝜕𝑦𝑖 ∗ 𝜕𝑦𝑖 𝜕𝑠𝑖 𝑂 ∗ 𝜕𝑠𝑖 𝑂 𝜕ℎ𝑗 𝑁 ∗ 𝜕ℎ𝑗 𝑁 𝜕𝑠𝑗 𝑁 ∗ 𝜕𝑠𝑗 𝑁 𝜕𝑤𝑗,𝑖 𝑁 = 𝑖 𝛿𝑖 𝑁 𝑤𝑖,𝑗 𝑁+1 𝑓′ 𝑠𝑗 𝑁 ℎ 𝑘 𝑁−1 (𝑁 − 𝛿𝑗 𝑁−1 = 𝑖 𝛿𝑖 𝑁 𝑤𝑖,𝑗 𝑁+1 𝑓′ 𝑠𝑗 𝑁 𝜕𝐿 𝜕𝑤𝑙,𝑘 𝑁−1 = 𝑗 𝛿𝑗 𝑁−1 𝑤𝑗,𝑘 𝑁 𝑓′ 𝑠 𝑘 𝑁−1 ℎ𝑙 𝑁−2 … 1 𝑠𝑡 hidden 𝛿 𝑘 1 𝜕𝐿 𝜕𝑤𝑖,𝑗 1 = 𝑘 𝛿 𝑘 1 𝑤 𝑘,𝑙 2 𝑓′ 𝑠𝑗 1 𝑥𝑖
  • 10. Challenges of training deep networks • Saturation • Vanishing gradients • Overfitting • Slowness of second order methods • Slow convergence, stucks in local optima with first order methods • (Exploding gradients)
  • 12. Breakthroughs in research • Saturation & vanishing gradients  Layer-by-layer training (2006)  Non-saturating activation functions, e.g. ReLU (2013) • Overfitting  Dropout (2014) • Convergence problems  Adagrad, Adadelta, Adam, RMSProp, etc.
  • 13. Computational power • Natural increase in computational power • GP GPU technology
  • 15. Don’t give in to the HYPE • Deep learning is impressive but  deep learning is not true AI o it may be a component of it when and if AI is created  deep learning is not how the human brain works  95% of machine learning tasks don’t require deep learning  deep learning requires a lot of computational power • Deep learning is a tool  which is successful in certain, previously very challenging domains (speech recognition, computer vision, NLP, etc.)  that excels in pattern recognition You are here
  • 17. From the Netflix prize... • Netflix prize (2006-2009)  Gave a huge push to recommender systems research  Determined the direction of research for years  Task: o Some (User, Item, Rating) known triplets o (User, Item) pairs with unknown rating o Predict the missing ratings (1-5)
  • 18. ... to recommenders in practice • Ratings  events [implicit feedback]  Lots of services don’t allow for rating  Majority of users don’t rate  Monitored passively  preferences have to be infered • Rating prediction  ranking [top N recommendations]  All that matters is the relevancy of the top N items  Rating prediction is biased • User  session / situation [session-based / context-driven recommendation]  Users are not logged in, identification is unreliable  Accounts used by multiple users  Aim of the session (e.g. buy a good laptop)  Similar behavior of different users in a situation, different behavior of the same user in different situations
  • 19. Challenges in RecSys • Session modeling  Most of the algorithms are personalized  A few are item-to-item o Recommends similar items o Also used for session-based recommendations (industry de facto standard)  There are no good session based solutions • Incorporating factors that influence user clicks  Users click based on what they see o Title o Product image o Description  and on their knowledge of the product o Usually harder to model o Except when the product is content (e.g. music)
  • 20. Deep learning to the rescue – Session modeling • Recurrent Neural Networks (RNN)  Sequence modeling  Hidden state: next state is based on the previous hidden state and the current input  „Infinite” depth  More sophisticated versions: GRU, LSTM • Needs to be adapted to the recommendation task • GRU4Rec:  Session-parallel minibatch training for handling the large variance in session lengths  Sampling the output for reasonable training times, without losing much accuracy  Ranking loss for better item ranking • Results: 15-30% improvement over item-to-item recommendations 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 RSC15 VIDEO Recall@20 Item-kNN GRU4Rec 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 RSC15 VIDEO MRR@20 Item-kNN GRU4Rec
  • 21. Other uses of deep learning for recsys • Incorporating content directly  Music, images, video, text  User influencing aspects of the items  Direct content representation • Context-state modeling from sensory data  IoT devices  Lot of sensory data  Some missing and noise  Infer context state and recommend accordingly • Interactive recommenders using chatbots • Personalized content generation  Today’s news  Images in personalized style with personalized content • Etc...
  • 22. There is work to be done • DL + RecSys research: just started  Last year: o 0 long papers, 1 short paper and 1 poster that is loosely connected  This year: o 10+ submissions to RecSys in this topic o DLRS 2016 workshop @ RecSys • Open questions  (More) Application areas  Adaptations required for the recsys problem  Scalability  Best practices  ...
  • 23. Thanks for your attention!