Deep learning: the future of recommendations
22 likes15,153 views
The document discusses the advancements and challenges of deep learning, particularly in the context of recommendation systems, highlighting its applications in fields such as speech recognition and computer vision. It explains deep learning algorithms, their training processes, and various neural network architectures like feedforward networks and recurrent neural networks (RNNs). The text emphasizes the significance of deep learning in enhancing recommendation accuracy while acknowledging that it is not a panacea for AI or all machine learning tasks.
![... 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](https://image.slidesharecdn.com/deeplearningtalkstartupsafary20160421-160422150918/85/Deep-learning-the-future-of-recommendations-18-320.jpg)
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)
- 11. Why now?
- 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
- 14. Intermission
- 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
- 16. Deep learning for RecSys
- 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!