Every two weeks, we find the most interesting data science links from around the web and collect them in Data Science Briefings, the DataMiningApps newsletter. Subscribe now for free if you want to be the first to get up to speed on interesting resources.
- The Malicious Use of Artificial Intelligence
A report written by 26 authors from 14 institutions, spanning academia, civil society, and industry. - Manifesto for Data Practices
“We believe these values and principles, taken together, describe the most effective, ethical, and modern approach to data teamwork.” - Algorithmic Impact Assessments: Toward Accountable Automation in Public Agencies
In the coming months, NYC Mayor Bill de Blasio will announce a new task force on “Automated Decision Systems” — the first of its kind in the United States. The task force will recommend how each city agency should be accountable for using algorithms and other advanced computing techniques to make important decisions. - Roundup Of Machine Learning Forecasts And Market Estimates, 2018
From Forbes: Machine learning patents grew at a 34% Compound Annual Growth Rate (CAGR) between 2013 and 2017, the third-fastest growing category of all patents granted. - Gartner’s 2018 Take on Data Science Tools
… It now at least includes some real players. - Big Companies Are Embracing Analytics, But Most Still Don’t Have a Data-Driven Culture
Perhaps the best news in this survey is that companies continue to believe they are getting value from their big data and AI projects. 73% of respondents said they have already received measurable value from these initiatives. - The GANfather: The man who’s given machines the gift of imagination
By pitting neural networks against one another, Ian Goodfellow has created a powerful AI tool. Now he, and the rest of us, must face the consequences. - Inside Amazon’s AI Flywheel
How deep learning came to power Alexa, Amazon Web Services, and nearly every other division of the company. - Is “Murder by Machine Learning” the New “Death by PowerPoint”?
What makes artificial intelligence/machine learning (AI/ML) champions confident that their technologies will be immune to comparably counterproductive outcomes? They shouldn’t be so sure. - Humans may not always grasp why AIs act. Don’t panic
Humans are inscrutable too. Existing rules and regulations can apply to artificial intelligence. - An AI just beat top lawyers at their own game
A new study, conducted by legal AI platform LawGeex in consultation with law professors from Stanford University, Duke University School of Law, and University of Southern California, pitted twenty experienced lawyers against an AI trained to evaluate legal contracts. - From imitation to innovation: How China became a tech superpower
In China, change comes so quickly that the future can arrive before the past is fully stripped away. - China’s great leap forward in science
Chinese investment is paying off with serious advances in biotech, computing and space. Are they edging ahead of the west? - China and the US are bracing for an AI showdown—in the cloud
Alibaba, Amazon, and others are adding ever more capable AI services to their cloud platforms. - Meet the Chinese Finance Giant That’s Secretly an AI Company
The smartphone payments business Ant Financial is using computer vision, natural language processing, and mountains of data to reimagine banking, insurance, and more. - Why Self-Taught Artificial Intelligence Has Trouble With the Real World
The latest artificial intelligence systems start from zero knowledge of a game and grow to world-beating in a matter of hours. But researchers are struggling to apply these systems beyond the arcade. - Cloud TPU machine learning accelerators now available in beta
Cloud TPUs are available in beta on Google Cloud Platform (GCP) to help machine learning (ML) experts train and run their ML models more quickly. - Benchmarking Google’s new TPUv2
“Nine months after the initial announcement, Google last week finally released TPUv2 to early beta users on the Google Cloud Platform. At RiseML, we got our hands on them and ran a couple of quick benchmarks. We’d like to share our experience and preliminary results.” - Palantir deployed a predictive policing system in New Orleans that even city council members don’t know about
The company provided software to a secretive NOPD program that traced people’s ties to other gang members, outlined criminal histories, analyzed social media, and predicted the likelihood that individuals would commit violence or become a victim. - Mark Felt-Tipped
Decensoring redacted texts using a simple approach. Wondering whether this can be automated through AI. - Machine Learning Crash Course
Google’s fast-paced, practical introduction to machine learning. - TensorFlow 1.6.0 released
With support for CUDA 9.0 and cuDNN 7. - JupyterLab is Ready for Users
“We are proud to announce the beta release series of JupyterLab, the next-generation web-based interface for Project Jupyter.” - Queryparser, an Open Source Tool for Parsing and Analyzing SQL
Uber open sources Queryparser, a tool for parsing and analyzing SQL queries. - Propel: Machine learning for Javascript
Propel provides a GPU-backed numpy-like infrastructure for scientific computing in JavaScript. JavaScript is a fast, dynamic language which, we think, could act as an ideal workflow for scientific programmers of all sorts. - A Comparative Analysis of Top 6 BI and Data Visualization Tools in 2018
“All of the highlighted visualization platforms have their own pros and cons, but all of them can also make your data talk and show its hidden values.” - Image2Mesh: A Learning Framework for Single Image 3D Reconstruction (paper)
“One challenge that remains open in 3D deep learning is how to efficiently represent 3D data to feed deep networks. Recent works have relied on volumetric or point cloud representations, but such approaches suffer from a number of issues such as computational complexity, unordered data, and lack of finer geometry. This paper demonstrates that a mesh representation (i.e. vertices and faces to form polygonal surfaces) is able to capture fine-grained geometry for 3D reconstruction tasks.” - DeepPavlov: An open source library for building end-to-end dialog systems and training chatbots.
An open-source conversational AI library, built on TensorFlow and Keras, and designed for NLP and dialog systems research and implementation and evaluation of complex conversational systems. - Keras reimplementation of “One pixel attack for fooling deep neural networks”
How simple is it to cause a deep neural network to misclassify an image if we are only allowed to modify the color of one pixel and only see the prediction probability? Turns out it is very simple. In many cases, we can even cause the network to return any answer we want. - Analyzing the Gender Representation of 34,476 Comic Book Characters
Female characters appear in superhero comics less often than males — but when they are included, how are they depicted? - FastPhotoStyle from NVIDIA
“This code repository contains an implementation of our fast photorealistic style transfer algorithm. Given a content photo and a style photo, the code can transfer the style of the style photo to the content photo.” - Fonts for Complex Data
Retail displays, packaged goods, financial reports and apps all present readers with a dizzying array of data. Here are a few ways to make quick work of their long lists, tiny annotations, and mighty stacks of numbers. - New Maps Reveal Global Fishing’s ‘Vast Scope Of Exploitation Of The Ocean’
SkyTruth and its collaborators tracked most of the world’s fishing vessels through an entire year by monitoring radio transmissions that most vessels now emit automatically in order to avoid collisions with each other. - Building Production Recommendation Systems
A thorough post from Cloudera outlining how to setup a big data ready product recommendation system. - What Do Data Scientists Need to Know about Containerization?
As Little as Possible. - Voyages in sentence space
Imagine a sentence gradient between two sentences—not a story, but a smooth interpolation of meaning. - Do neural nets dream of electric sheep?
Image recognition algorithms can make really bizarre mistakes. - Google-Landmarks: A New Dataset and Challenge for Landmark Recognition
“Today, we are excited to advance instance-level recognition by releasing Google-Landmarks, the largest worldwide dataset for recognition of human-made and natural landmarks.” - GDPR – A practical guide for developers
“I will not go into yet another “12 facts about GDPR” or “7 myths about GDPR” posts/whitepapers, as they are often aimed at managers or legal people. Instead, I’ll focus on what GDPR means for developers.” - Hacking the Brain With Adversarial Images
Researchers from Google Brain show that adversarial images can trick both humans and computers, and the implications are scary. - AI Just Learned How to Boost the Brain’s Memory
By triggering precisely timed pulses of electricity to the brain, researchers can essentially use one black box to unlock the potential of another.
