Papers of the day   All papers

Deep and Confident Prediction for Time Series at Uber

Comments

Nov 08 2019 Xavier Amatriain

Time series prediction is clearly a very important problem for @UberEng. Here is a blogpost and a couple of papers from their presentation at #MLconfSF: https://eng.uber.com/neural-networks-uncertainty-estimation/ https://arxiv.org/abs/1709.01907 https://arxiv.org/abs/1906.04397
2 replies, 656 likes


Nov 09 2019 Emtiyaz Khan

Uber is using Bayesian NN which is great! I hope we can design better methods than MC dropout in the future.
0 replies, 17 likes


Jan 09 2020 HotComputerScience

Most popular computer science paper of the day: "Probabilistic Forecasting with Temporal Convolutional Neural Network" https://hotcomputerscience.com/paper/probabilistic-forecasting-with-temporal-convolutional-neural-network https://twitter.com/xamat/status/1192856676707078145
0 replies, 3 likes


Jan 13 2020 HotComputerScience

Most popular computer science paper of the day: "Deep and Confident Prediction for Time Series at Uber" https://hotcomputerscience.com/paper/deep-and-confident-prediction-for-time-series-at-uber https://twitter.com/xamat/status/1192856676707078145
0 replies, 1 likes


Content