Скачать с ютуб Why Deep Neural Networks (DNNs) Underperform Tree-Based Models on Tabular Data в хорошем качестве

Why Deep Neural Networks (DNNs) Underperform Tree-Based Models on Tabular Data

In this video we talk about why deep neural networks (DNNs) fail to beat tree-based models like random forest or xgboost on tabular datasets. We analyse, step by step, the arguments invoked in the paper "Why do tree-based models still outperform deep learning on tabular data?". Related Videos ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ Why Neural Networks Can Learn Any Function:    • Why Neural Networks Can Learn Any Fun...   Why neural networks are deep:    • Why Neural Networks (NN) Are Deep | T...   Why residual connections work:    • Why Residual Connections (ResNet) Work   Why we need activations in neural nets:    • Why We Need Activation Functions In N...   References ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ Why do tree-based models still outperform deep learning on tabular data?: https://arxiv.org/abs/2207.08815 Contents ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 00:00 - Intro 01:00 - Argument 1 - Decision Boundary and Irregular Patterns 02:37 - Argument 2 - Uninformative Features 03:49 - Argument 3 - Rotation Invariance 04:46 - Summary 05:23 - Outro Follow Me ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 🐦 Twitter: @datamlistic   / datamlistic   📸 Instagram: @datamlistic   / datamlistic   📱 TikTok: @datamlistic   / datamlistic   Channel Support ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ The best way to support the channel is to share the content. ;) If you'd like to also support the channel financially, donating the price of a coffee is always warmly welcomed! (completely optional and voluntary) ► Patreon:   / datamlistic   ► Bitcoin (BTC): 3C6Pkzyb5CjAUYrJxmpCaaNPVRgRVxxyTq ► Ethereum (ETH): 0x9Ac4eB94386C3e02b96599C05B7a8C71773c9281 ► Cardano (ADA): addr1v95rfxlslfzkvd8sr3exkh7st4qmgj4ywf5zcaxgqgdyunsj5juw5 ► Tether (USDT): 0xeC261d9b2EE4B6997a6a424067af165BAA4afE1a #neuralnetworks #tabulardata #structureddata