MLconf SF 2016
2016-11-11
San Francisco, United States
[MLconf SF 2014, MLconf SF 2015, MLconf SF 2016, MLconf SF 2017, MLconf SF 2019]
| title | author | length |
|---|---|---|
| Machine Learning with TensorFlow | Rajat Monga | 26m |
TensorFlow is an open C++ core with Python and other front ends that runs on CPU GPU and TPU and claims 1,000,000+ downloads and nearly 500 contributors (over… | ||
| Neural Turing Machines: Perils and Promise | Daniel Shank | 21m |
Neural Turing Machines (NTMs) are differentiable controllers plus external memory that learn algorithms like copying and balanced-parentheses; implemented in… | ||
| Using Deep Reinforcement Learning for Dialogue Systems | Harm van Seijen | 21m |
Describes training a dialogue policy with deep reinforcement learning using DQN variants and an actor‑critic baseline, adding experience replay and a slowly… | ||
| Being Smart with Art | Guy Lebanon | 27m |
Netflix uses machine learning and A/B explore‑exploit to pick and serve artwork, claiming measurable lifts in streaming hours and valuing the recommender at… | ||
| Before the Model: How Machine Learning Products Start... | Elena Grewal | 22m |
Describes how Airbnb turns vague ideas into deployed ML products by sizing impact, choosing a business metric, and reformulating problems — e.g., a six‑month… | ||
| Amazon Search: The Joy of Ranking Products | Daria Sorokina | 21m |
Amazon Search: The Joy of Ranking Products: Amazon is one of the world’s largest e-commerce sites and Amazon Search powers the majority of Amazon’s sales. As a… | ||
| Personalized Content Blending In The Pinterest Homefeed | Stephanie deWet | 19m |
Pinterest builds a homefeed by ranking separate content pools (followed users, followed topics, recommendations) and blends them with a personalized integer… | ||
| A General Framework for Communication-Efficient Distributed... | Virginia Smith | 15m |
Proposes the CoCoA framework that plugs existing single‑machine solvers into a primal‑dual distributed loop, applies local updates immediately, and averages… | ||
| Interpreting Black-Box Models with Applications to Healthcare | Brian Lucena | 42m |
Toolkit pre-computes model outputs by mutating each feature across its range and produces feature dependence plots, a feature effect summary that ranks… | ||
| Personalization and Scalable Deep Learning with | Alex Smola | 43m |
Claims that time-varying personalization yields better results when modeled with an LSTM-based recurrent latent-variable model that predicts user return times… | ||
| Smart Reply: Learning a Model of Conversation from Data | Anjuli Kannan | 24m |
Smart Reply generates one-tap responses for Inbox by Gmail by learning directly from data: a sequence-to-sequence recurrent neural network encodes incoming… | ||
| A Friendly Introduction To Causality | Alex Dimakis | 24m |
Surveys Pearl’s Structural Equation Models and shows interventions are the gold standard: randomize a variable (example: 50% treatment vs placebo) and check… | ||
| Using Bayesian Optimization to Tune Machine Learning Models | Scott Clark | 24m |
Argues that Bayesian global optimization (Gaussian-process surrogate with expected improvement acquisition) finds better hyperparameters with far fewer costly… | ||
| Several People Are Tuning: Data and Machine Learning at | Josh Wills | 29m |
Slack has two years of public data, a newly formed machine learning team (six months) and a seven‑person data engineering team that builds a modern stack on… | ||
| Why Machine Learning Algorithms Fall Short... | Jean-François Puget | 26m |
Argues that production ML needs more than algorithms: IBM Watson Machine Learning beta serializes models to JSON and ships a separate efficient serving engine… | ||
| Building a Machine Learning Platform at | Nikhil Garg | 18m |
Presentation slides: http://www.slideshare.net/SessionsEvents/nikhil-garg-engineering-manager-quora-at-mlconf-sf-2016
Building a Machine Learning Platform at… | ||
| Review Analysis: an Approach to Leveraging User-Generated... | Jennifer Prendki | 26m |
Pipeline extracts product-specific decision features via tf-idf + word2vec and an entity graph, runs spellchecking and POS filters, then scores per-feature… | ||
| Q&A Session | Rajat Monga, Nikolaos Vasiloglou | 18m |
Answers cover TensorFlow behavior and tradeoffs: single‑machine task and data parallelism works out‑of‑the‑box for typical cases; small matrices favor NumPy… | ||
| Tianqi Chen, Industry Impact Student Research Award Winner | Tianqi Chen | 01m |
Describes building scalable systems that bring tree boosting and deep learning to industry-scale data, combining distributed computation with out-of-core… | ||
| nothing matches | ||