TMLS2018 - Machine Learning in Production, Panel Discussion

Channel:
Canada LLMs Explained - Aggregate Intellect - AI.SCIENCE
Subscribers:
22,300
Published on ● Video Link: https://www.youtube.com/watch?v=s9vygTVYegc


Category:
Discussion
Duration: 1:07:06
1,733 views
22

Part of Toronto Machine Learning (micro) Summit 2018. The panelists discussed the nuances of "machine learning in production".

Hosted by: Amir Feizpour (http://amirfeizpour.pythonanywhere.com/)

Amir Hajian (https://www.linkedin.com/in/amir-hajian-744674135/)
Vincent Wong (https://www.linkedin.com/in/vincent-f-wong-289478121/)
Solmaz Shahalizadeh (https://www.linkedin.com/in/solmazshahalizadeh/)
Inmar Givoni (https://www.linkedin.com/in/inmargivoni/)

Questions:
1:30
Introduce the panelists, and why they're interested in this topic

8:40
Solmaz, your team deploys a lot of machine learning models into production. Can you talk a little about that process, and how it is different from the exploration phase?

12:20
Amir, I have heard you talk about the importance of design in ML products a few times. And that’s a very important issue, because not only you have to design your product with the user in mind, but deploying a model is just the beginning of getting a lot of interesting feedback from the user or the system it is integrating into. Can you walk us through your thoughts there?

17:23
"Minimum viable machinery"

23:30
How would you create intimate relationships between data scientists and production team, and users to get their feedback?

30:56
Inmar, you are working for company that grew very quickly and must be a very fast paced environment. In general, and in the case of your company specifically, I think scale is a very important issue. Can you speak about the importance of that and how you manage it?

36:15
Can data engineers who are responsible for bringing the data the same ones as those who do productionization? or do you suggest to use different teams for that?

40:18
Vincent, you are part of a company that integrates machine learning in the already existing platforms and infrastructure. Can you talk about the nuances of that?

45:18
Nuances of interacting with other IT team and clients

48:06
Should data scientists write code given their lack of software development skills?

50:18
Let’s wrap up by hearing your advice for people who are just starting to think about the importance of ML in production. What’s the most important thing that you think they should think very carefully about?



Other Videos By LLMs Explained - Aggregate Intellect - AI.SCIENCE


2018-09-24Junction Tree Variational Autoencoder for Molecular Graph Generation | TDLS
2018-09-19Reconstructing quantum states with generative models | TDLS Author Speaking
2018-09-13All-optical machine learning using diffractive deep neural networks | TDLS
2018-09-05Recurrent Models of Visual Attention | TDLS
2018-08-28Eve: A Gradient Based Optimization Method with Locally and Globally Adaptive Learning Rates | TDLS
2018-08-20TDLS: Large-Scale Unsupervised Deep Representation Learning for Brain Structure
2018-08-14Principles of Riemannian Geometry in Neural Networks | TDLS
2018-08-07Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond | TDLS
2018-07-30Program Language Translation Using a Grammar-Driven Tree-to-Tree Model | TDLS
2018-07-23Explainable Neural Networks based on Additive Index Models | TDLS
2018-07-18TMLS2018 - Machine Learning in Production, Panel Discussion
2018-07-16Flexible Neural Representation for Physics Prediction | AISC Trending Paper
2018-07-10Connectionist Temporal Classification, Labelling Unsegmented Sequence Data with RNN | TDLS
2018-06-25Learning to Represent Programs with Graphs | TDLS
2018-06-19Quantum generative adversarial networks | TDLS Author Speaking
2018-06-12[SAGAN] Self-Attention Generative Adversarial Networks | TDLS
2018-06-05[ELMo] Deep Contextualized Word Representations | AISC
2018-05-23Few-Shot Learning Through an Information Retrieval Lens | TDLS
2018-05-14Improving Supervised Bilingual Mapping of Word Embeddings | TDLS
2018-05-01[Word2Bits] Quantized Word Vectors | AISC
2017-11-15TMLS2017: Transitioning to Data Science, Panel Discussion


Tags:
machine learning
AI devops
panel discussion
discussion panel
ML-ops
machine learning production