Deep Networks Are Kernel Machines (Paper Explained)
#deeplearning #kernels #neuralnetworks
Full Title: Every Model Learned by Gradient Descent Is Approximately a Kernel Machine
Deep Neural Networks are often said to discover useful representations of the data. However, this paper challenges this prevailing view and suggest that rather than representing the data, deep neural networks store superpositions of the training data in their weights and act as kernel machines at inference time. This is a theoretical paper with a main theorem and an understandable proof and the result leads to many interesting implications for the field.
OUTLINE:
0:00 - Intro & Outline
4:50 - What is a Kernel Machine?
10:25 - Kernel Machines vs Gradient Descent
12:40 - Tangent Kernels
22:45 - Path Kernels
25:00 - Main Theorem
28:50 - Proof of the Main Theorem
39:10 - Implications & My Comments
Paper: https://arxiv.org/abs/2012.00152
Street Talk about Kernels: https://youtu.be/y_RjsDHl5Y4
ERRATA: I simplify a bit too much when I pit kernel methods against gradient descent. Of course, you can even learn kernel machines using GD, they're not mutually exclusive. And it's also not true that you "don't need a model" in kernel machines, as it usually still contains learned parameters.
Abstract:
Deep learning's successes are often attributed to its ability to automatically discover new representations of the data, rather than relying on handcrafted features like other learning methods. We show, however, that deep networks learned by the standard gradient descent algorithm are in fact mathematically approximately equivalent to kernel machines, a learning method that simply memorizes the data and uses it directly for prediction via a similarity function (the kernel). This greatly enhances the interpretability of deep network weights, by elucidating that they are effectively a superposition of the training examples. The network architecture incorporates knowledge of the target function into the kernel. This improved understanding should lead to better learning algorithms.
Authors: Pedro Domingos
Links:
TabNine Code Completion (Referral): http://bit.ly/tabnine-yannick
YouTube: https://www.youtube.com/c/yannickilcher
Twitter: https://twitter.com/ykilcher
Discord: https://discord.gg/4H8xxDF
BitChute: https://www.bitchute.com/channel/yannic-kilcher
Minds: https://www.minds.com/ykilcher
Parler: https://parler.com/profile/YannicKilcher
LinkedIn: https://www.linkedin.com/in/yannic-kilcher-488534136/
BiliBili: https://space.bilibili.com/1824646584
If you want to support me, the best thing to do is to share out the content :)
If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this):
SubscribeStar: https://www.subscribestar.com/yannickilcher
Patreon: https://www.patreon.com/yannickilcher
Bitcoin (BTC): bc1q49lsw3q325tr58ygf8sudx2dqfguclvngvy2cq
Ethereum (ETH): 0x7ad3513E3B8f66799f507Aa7874b1B0eBC7F85e2
Litecoin (LTC): LQW2TRyKYetVC8WjFkhpPhtpbDM4Vw7r9m
Monero (XMR): 4ACL8AGrEo5hAir8A9CeVrW8pEauWvnp1WnSDZxW7tziCDLhZAGsgzhRQABDnFy8yuM9fWJDviJPHKRjV4FWt19CJZN9D4n
Видео Deep Networks Are Kernel Machines (Paper Explained) канала Yannic Kilcher
Full Title: Every Model Learned by Gradient Descent Is Approximately a Kernel Machine
Deep Neural Networks are often said to discover useful representations of the data. However, this paper challenges this prevailing view and suggest that rather than representing the data, deep neural networks store superpositions of the training data in their weights and act as kernel machines at inference time. This is a theoretical paper with a main theorem and an understandable proof and the result leads to many interesting implications for the field.
OUTLINE:
0:00 - Intro & Outline
4:50 - What is a Kernel Machine?
10:25 - Kernel Machines vs Gradient Descent
12:40 - Tangent Kernels
22:45 - Path Kernels
25:00 - Main Theorem
28:50 - Proof of the Main Theorem
39:10 - Implications & My Comments
Paper: https://arxiv.org/abs/2012.00152
Street Talk about Kernels: https://youtu.be/y_RjsDHl5Y4
ERRATA: I simplify a bit too much when I pit kernel methods against gradient descent. Of course, you can even learn kernel machines using GD, they're not mutually exclusive. And it's also not true that you "don't need a model" in kernel machines, as it usually still contains learned parameters.
Abstract:
Deep learning's successes are often attributed to its ability to automatically discover new representations of the data, rather than relying on handcrafted features like other learning methods. We show, however, that deep networks learned by the standard gradient descent algorithm are in fact mathematically approximately equivalent to kernel machines, a learning method that simply memorizes the data and uses it directly for prediction via a similarity function (the kernel). This greatly enhances the interpretability of deep network weights, by elucidating that they are effectively a superposition of the training examples. The network architecture incorporates knowledge of the target function into the kernel. This improved understanding should lead to better learning algorithms.
Authors: Pedro Domingos
Links:
TabNine Code Completion (Referral): http://bit.ly/tabnine-yannick
YouTube: https://www.youtube.com/c/yannickilcher
Twitter: https://twitter.com/ykilcher
Discord: https://discord.gg/4H8xxDF
BitChute: https://www.bitchute.com/channel/yannic-kilcher
Minds: https://www.minds.com/ykilcher
Parler: https://parler.com/profile/YannicKilcher
LinkedIn: https://www.linkedin.com/in/yannic-kilcher-488534136/
BiliBili: https://space.bilibili.com/1824646584
If you want to support me, the best thing to do is to share out the content :)
If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this):
SubscribeStar: https://www.subscribestar.com/yannickilcher
Patreon: https://www.patreon.com/yannickilcher
Bitcoin (BTC): bc1q49lsw3q325tr58ygf8sudx2dqfguclvngvy2cq
Ethereum (ETH): 0x7ad3513E3B8f66799f507Aa7874b1B0eBC7F85e2
Litecoin (LTC): LQW2TRyKYetVC8WjFkhpPhtpbDM4Vw7r9m
Monero (XMR): 4ACL8AGrEo5hAir8A9CeVrW8pEauWvnp1WnSDZxW7tziCDLhZAGsgzhRQABDnFy8yuM9fWJDviJPHKRjV4FWt19CJZN9D4n
Видео Deep Networks Are Kernel Machines (Paper Explained) канала Yannic Kilcher
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
ReBeL - Combining Deep Reinforcement Learning and Search for Imperfect-Information Games (Explained)Deep Learning: A Crash CourseThis equation will change how you see the world (the logistic map)The Symmetries of the universeHistory of Bayesian Neural Networks (Keynote talk)The Kernel Trick - THE MATH YOU SHOULD KNOW!Kernels!Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate ShiftHow Does Gravity Warp the Flow of Time?Feedback Transformers: Addressing Some Limitations of Transformers with Feedback Memory (Explained)The AI Economist: Improving Equality and Productivity with AI-Driven Tax Policies (Paper Explained)Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention (AI Paper Explained)OpenAI CLIP: ConnectingText and Images (Paper Explained)The Electric Vehicle Charging ProblemNeural Network that Changes Everything - ComputerphileBayes theoremMIT Introduction to Deep Learning | 6.S191Big Self-Supervised Models are Strong Semi-Supervised Learners (Paper Explained)Zero Knowledge Proof - Numberphile16. Learning: Support Vector Machines