Thinking about Machine Learning Data Pipelines

cyberpunk

Almost every day I have a new idea of some machine learning model for an app or some feature. Machine learning involves tasks that include data sourcing, data ingestion, data transformation, pre-processing data for use in training, training a model and hosting the model. Additionally, to get value out of ...

more ...

Studies in AI & Pixels & Waves - #12

Sat 08 October 2016 Category AI & ML

Studies in Machine Learning & AI - #11

Sun 25 September 2016 Category AI & ML

Studies in Machine Learning & AI - #10

Sat 17 September 2016 Category AI & ML

Studies in Machine Learning & AI - #9

Sat 10 September 2016 Category AI & ML

Studies in Machine Learning & AI - #8

Sat 03 September 2016 Category AI & ML

Studies in Machine Learning & AI - #7

Sat 27 August 2016 Category AI & ML

cyber

Papers

  • Decoupled Neural Interfaces using Synthetic Gradients. New DeepMind's paper reviewing backpropagation, using a modeled synthetic gradient in place of true backpropagated error gradients. Backpropagation is a bottleneck in DNN, so the idea is instead to use an asynchronous estimator, that would be obtained by supervised training of another ...
more ...

Studies in Machine Learning & AI - #6

Fri 19 August 2016 Category AI & ML

Studies in Machine Learning & AI - #5

Sun 14 August 2016 Category AI & ML

Studies in Machine Learning & AI - #4

Sat 06 August 2016 Category AI & ML

cyberpunk

Papers

News

  • Costs of extinction risk mitigation. A Cost-Benefit Analysis of the extinction risk mitigation, claiming that the annual cost of reducing the probability of human extinction by 0.01 ...
more ...