DataFramed

Hugo speaks with Peter Bull about the importance of human-centered design in data science. Peter is a data scientist for social good and co-founder of Driven Data, a company that brings cutting-edge practices in data science and crowdsourcing to some of the world's biggest social challenges and the organizations taking them on, including machine learning competitions for social good. They’ll speak about the practice of considering how humans interact with data and data products and how important it is to consider them while designing your data projects. They’ll see how human-centered design provides a robust and reproducible framework for involving the end-user all through the data work, illuminated by examples such as DrivenData’s work in financial services and Mobile Money in Tanzania. Along the way, they’ll discuss the role of empathy in data science, the increasingly important conversation around data ethics and much, much more.

LINKS FROM THE SHOW

FROM THE INTERVIEW

• Peter on Twitter
• DrivenData
• Deon (Ethics Checklist)
• Cookiecutter Data Science
• If you liked this interview, you might be interested in working with DrivenData! Currently, the team is looking for a software engineer who loves the idea of building Python applications for social impact. Apply Here!

FROM THE SEGMENTS

Probability Distributions and their Stories (with Justin Bois at ~24:00)

• Justin's Website at Caltech
• Probability distributions and their stories (By Justin Bois)

Studies in Interpretability (with Peadar Coyle at ~38:10)

• Interpretable ML Symposium
• How will the GDPR impact machine learning? (By Andrew Burt)
• How to use Bayesian Stats in your daily job (Gates, Perry, Zorn (2002))
• Fairness in Machine Learning (By Moritz Hardt)

Original music and sounds by The Sticks.

In this episode of DataFramed, a DataCamp podcast, Hugo speaks with Arnaub Chatterjee. Arnaub is a Senior Expert and Associate Partner in the Pharmaceutical and Medical Products group at McKinsey & Company. They’ll discuss cutting through the hype about artificial intelligence (AI) and machine learning (ML) in healthcare by looking at practical applications and how McKinsey & Company is helping the industry evolve.

Tune in for an insider’s account into what has worked in healthcare, from ML models being used to predict nearly everything in clinical settings, to imaging analytics for disease diagnosis, to wound therapeutics. Will robots and AI replace disciplines such as radiology, ophthalmology, and dermatology? How have the moving parts of data science work evolved in healthcare? What does the future of data science, ML and AI in healthcare hold? Stick around to find out.

LINKS FROM THE SHOW

FROM THE INTERVIEW

• McKinsey Analytics on Twitter
• Hot off the press article for HBR’s Future of Healthcare online forum (By Arnaub Chatterjee)
• Our latest piece on the promise & challenge of AI (By James Manyika and Jacques Bughin)
• Are robots coming for our jobs? (mckinsey.com)
• Analytics Careers page (mckinsey.com)
• How we help clients in healthcare analytics (mckinsey.com)
• AI analysis of 400+ use cases, including ones in healthcare (By Michael Chui et al. mckinsey.com)

FROM THE SEGMENTS

Machines that Multi-task (with Manny Moss)

Part 1 at ~21:05

• Responsible AI in Consumer Enterprise
• Hilary Mason, DJ Patil and Mike Loukides on Data Ethics
• EthicalOS Tookit

Part 2 at ~40:00

• 21 Definitions of Fairness Tutorial from FAT* (Arvind Naranayan)
• Kate Crawford's keynote address "The Trouble with Bias" from NIPS 2017
• The (im)possibility of Fairness (Sorelle et al. arXiv.org)
• Learning from disparate data sources (Li Y et al. PubMed.gov)
• Distributed Multi-task Learning (Liyang Xie et al. KDD.org)
• The Cost of Fairness in Binary Classification (Aditya Krishna Menon et al. proceedings.mlr.press)

Original music and sounds by The Sticks.

In this episode of DataFramed, Hugo speaks with Cassie Kozyrkov, Chief Decision Scientist at Google Cloud. Cassie and Hugo will be talking about data science, decision making and decision intelligence, which Cassie thinks of as data science plus plus, augmented with the social and managerial sciences. They’ll talk about the different and evolving models for how the fruits of data science work can be used to inform robust decision making, along with pros and cons of all the models for embedding data scientists in organizations relative to the decision function. They’ll tackle head on why so many organizations fail at using data to robustly inform decision making, along with best practices for working with data, such as not verifying your results on the data that inspired your models. As Cassie says, “Split your damn data”.

Links from the show

FROM THE INTERVIEW

• Cassie on Twitter 
• Is data science a bubble? (By Cassie Kozyrkov, Hackernoon)
• Incompetence, delegation, and population (By Cassie Kozyrkov, Hackernoon)
• Populations — You’re doing it wrong (By Cassie Kozyrkov, Hackernoon)
• What on earth is data science? (By Cassie Kozyrkov, Hackernoon)

FROM THE SEGMENTS

Probability Distributions and their Stories (with Justin Bois at ~19:45)

• Justin's Website at Caltech
• Probability distributions and their stories (By Justin Bois)

Machines that Multi-Task (with Friederike Schüür of Fast Forward Labs ~43:45)

• Sebastian’s Ruder’s Overview of Multi-Task Learning in Deep Neural Networks
• Multi-Task Learning for NLP, also by Sebastian Ruder
• GANs for Fake Celebrity Images (Karras et al, Nvidia)
• Adversarial Multi-Task Learning for Text Classification (Liu et al., arXiv.org)

Original music and sounds by The Sticks.

In this episode of DataFramed, Hugo speaks with Brian Granger, co-founder and co-lead of Project Jupyter, physicist and co-creator of the Altair package for statistical visualization in Python.

They’ll speak about data science, interactive computing, open source software and Project Jupyter. With over 2.5 million public Jupyter notebooks on github alone, Project Jupyter is a force to be reckoned with. What is interactive computing and why is it important for data science work? What are all the the moving parts of the Jupyter ecosystem, from notebooks to JupyterLab to JupyterHub and binder and why are they so relevant as more and more institutions adopt open source software for interactive computing and data science? From Netflix running around 100,000 Jupyter notebook batch jobs a day to LIGO’s Nobel prize winning discovery of gravitational waves publishing all their results reproducibly using Notebooks, Project Jupyter is everywhere. 

Links from the show 

FROM THE INTERVIEW

• Brian on Twitter 
• Project Jupyter
• Beyond Interactive: Notebook Innovation at Netflix (Ufford, Pacer, Seal, Kelley, Netflix Tech Blog)
• Gravitational Wave Open Science Center (Tutorials)
• JupyterCon YouTube Playlist
• jupyterstream Github Repository

FROM THE SEGMENTS

Machines that Multi-Task (with Friederike Schüür of Fast Forward Labs)

Part 1 at ~24:40

• Brief Introduction to Multi-Task Learning (By Friederike Schüür)
• Overview of Multi-Task Learning Use Cases (By Manny Moss)
• Multi-Task Learning for the Segmentation of Building Footprints (Bischke et al., arXiv.org)
• Multi-Task as Question Answering (McCann et al., arXiv.org)
• The Salesforce Natural Language Decathlon: A Multitask Challenge for NLP 

Part 2 at ~44:00

• Rich Caruana’s Awesome Overview of Multi-Task Learning and Why It Works
• Sebastian’s Ruder’s Overview of Multi-Task Learning in Deep Neural Networks
• Massively Multi-Task Network for Drug Discovery, 259 Tasks (!) (Ramsundar et al. arXiv.org)
• Brief Overview of Multi-Task Learning with Video of Newsie, the Prototype (By Friederike Schüür)

 Original music and sounds by The Sticks.

Hugo speaks with Andrew Gelman about statistics, data science, polling, and election forecasting. Andy is a professor of statistics and political science and director of the Applied Statistics Center at Columbia University and this week we’ll be talking the ins and outs of general polling and election forecasting, the biggest challenges in gauging public opinion, the ever-present challenge of getting representative samples in order to model the world and the types of corrections statisticians can and do perform. 

"Chatting with Andy was an absolute delight and I cannot wait to share it with you!"-Hugo 

 Links from the show 

FROM THE INTERVIEW

• Andrew's Blog 
• Andrew on Twitter 
• We Need to Move Beyond Election-Focused Polling (Gelman and Rothschild, Slate)
• We Gave Four Good Pollsters the Same Raw Data. They Had Four Different Results (Cohn, The New York Times).
• 19 things we learned from the 2016 election (Gelman and Azari, Science, 2017)
• The best books on How Americans Vote (Gelman, Five Books)
• The best books on Statistics (Gelman, Five Books)
• Andrew's Research 

FROM THE SEGMENTS

Statistical Lesson of the Week (with Emily Robinson at ~13:30)

• The five Cs (Loukides, Mason, and Patil, O'Reilly)

Data Science Best Practices (with Ben Skrainka~40:40)

• Oberkampf & Roy’s Verification and Validation in Scientific Computing provides a thorough yet very readable treatment 
• A comprehensive framework for verification, validation, and uncertainty quantification in scientific computing (Roy and Oberkampf, Science Direct)

 Original music and sounds by The Sticks.