MACHINE LEARNING

Top 50: Technology

The O'Reilly Data Show explores the opportunities and techniques driving big data and data science. Through interviews and analysis, we highlight the people putting data to work.

The O'Reilly Data Show Podcast

Machine learning

Lyt med, når IDA Podcast sætter fokus på evolutionære algoritmer og kunstige neurale netværk. Det er kognitive teknologier, som er fuldstændigt centrale for udviklingen indenfor kunstig intelligens i dag - og ikke til at overse, når vi skal forstå det eksponentielt accelererende paradigme, vi lever i. Vi stiller også skarpt på spørgsmålet, om maskiner kan være kreative, og på den vigtige rolle, kunstig intelligens spiller i en verden, hvor datamængden og processorkraften bare vokser og vokser. 

Podcasten er produceret af Danmarks ingeniørforening, IDA, i samarbejde med Brain Gain Group. Episoden er den tredje i en serie om fremtidsteknologi.

Medvirkende
Sebastian Risi, Associate Professor ved ITU og co-director for forskningsenheden Robotics, Evolution and Art Laboratory (REAL): http://bit.ly/2kTEKOC 

Thomas Terney, ph.d. i kunstig intelligens, foredragsholder og iværksætter: http://bit.ly/2kOQlik

Vært og tilrettelæggelse: Matias Seidler
Producer: Tobias Ankjær Jeppesen
Lyddesign: Alexander Clerici



SHOW NOTES

[00:23] IBMs Deep Blue blev den første computer som slog en stormester, Garry Kasparov, i skak. Det skete i 1997: http://bit.ly/2kIkvkB

[01:32] Link til præsentation af Henry Lieberman, MIT Media Lab, som forklarer forskellen mellem symbolsk (klassisk) kunstig intelligens og subsymbolsk kunstig intelligens: http://bit.ly/2k4lN7X

[02:05] For en frisk introduktion til Deep Learning, tjek WIREDs artikel, ‘Why We Need To Tame Our Algorithms Like Dogs’: http://bit.ly/2kIIqQV

[02:45] Forskningsenheden på ITU, ‘Robotics, Evolution and Arts Lab’, kan du kigge nærmere på her: http://bit.ly/2ksAxAE

[03:27] For en illustrativ oversigt over hvad biologisk inspirerede algoritmer har af anvendelsesmuligheder kan du tjekke tag-søgningen på Robohub.org ud: http://bit.ly/2knkZf1

[04:42] Der er flere internationale priser og konkurrencer om at løse ‘General Artificial Intelligence’-udfordringen. Se bl.a. denne på 35M$: http://bit.ly/2knrUVA

[06:32] Thomas Terney taler om styrkeforholdet mellem neuroner (eller enheder i netværket). Det betegnes som ‘weights’ på engelsk. Her er der en tråd med en række forskelligartede, uddybende besvarelser: http://bit.ly/2k40vMx 

[08:51] Baidu er den kinesiske ækvivalent til Google og har i januar 2017 hyret nogle af Microsofts dygtigste AI-udviklere: http://bit.ly/2lrPuS1

[10:34] Se NASAs whitepaper ‘Automated Antenna Design with Evolutionary Algorithms’: http://go.nasa.gov/2llsiYO
 
[11:32] MIT Technology Review har en interessant artikel med løsningsforslag på problematikken: ‘Algorithms That Learn with Less Data Could Expand AI’s Power’: http://bit.ly/2ksG1M2


[14:23] I 2016 trak Googles AlphaGo headlines i hele verden da den slog Lee Sedol i det - i forhold til skak usammenligneligt - komplekse spil, ‘Go’. Og AlphaGo bliver ved med at sejre, se bare her: http://bit.ly/2k4t7jK

[19:57] Apple er i fuld gang med at anvende ‘unsupervised learning’-princippet i udviklingen af selvkørende biler: http://bit.ly/2lplN3e

[21:23] ‘Do you think computers have minds?’ Det er ikke kun Sebastian Risi, som synes det spørgsmål er godt. Bevidsthedsfilosofien har søgt seriøse svar på det, helt siden Alan Turing formulerede sin berømte test i 1950. Det er blevet omdrejningspunktet for en filosofisk tradition, hvis mest interessante og udfoldede svar elegant beskrives i The Internet Encyclopedia of Philosophy: http://bit.ly/2k4Bqw5

Machine Learning

Deep Learning - New Age for Study, Improve Concentration and Brain Power, Instrumental Study Music for Exam Study, to Focus on Learning

Deep Learning

Deep Learning – Instrumental Relaxing Music for Reading, Piano & Flute Sounds to Increase Brain Power, New Age Concentration Music

Fraud attacks have become much more sophisticated. Account takeovers are happening more often. Many security attacks involve multiple methods and unexpected attacks can devastate businesses in just a few days, as we saw with Neiman Marcus and Target. False promotion and abuse is seen not only on social media sites but is also targeted at business. To combat these risks, fraud solutions need to be smarter to keep pace with fraudsters to prevent attacks and react quickly when they do happen. This requires a fast-learning solution with the ability to continually evolve. In this episode we talk to Kevin Lee from Sift Science and examine the shifts in the info security landscape over the past ten or fifteen year. Lee also highlights what new kinds of fraud are now possible and what machine learning solutions are available. See more episodes at: www.TechEmergence.com &nbsp;

Machine Learning for Fraud Detection - Modern Applications and Risks

# Invitado: Federico Baylé

- Emprendimiento imagenes satelitales: http://dymaxionlabs.com/ 
- El mapa de las villas/asentamientos: http://dymaxionlabs.com/potencialesvya/ 
- Escuela data science, dónde da clases: http://eant.tech/escuela-de-ciencias-de-datos
- Su maestría UBA: http://datamining.dc.uba.ar/datamining/
- Su cuenta de twitter: https://twitter.com/federicobayle

# foto con boudou

https://twitter.com/fernandezpablo/status/926450557266661376

# rant de crosa sobre IA/ML

https://medium.com/@alejandrocrosa/dejen-de-vender-humo-diciendo-inteligencia-artificial-61ba0eb50b46

# distintos tipos de redes neuronales

http://www.asimovinstitute.org/neural-network-zoo/

# the botanic app

http://botanicapp.com/

# CoreML

https://developer.apple.com/documentation/coreml

# tensorflow, scikit-learn, kreas

https://www.tensorflow.org/
http://scikit-learn.org/
https://keras.io/

# Siraj Raval (youtube)

https://www.youtube.com/channel/UCWN3xxRkmTPmbKwht9FuE5A

Three themes from Detroit: Become Human, each focusing on one of the title's main characters, and each scored by a different composer.

Detroit: Become Human launches May 25, 2018 on PS4.

Detroit: Become Human — Markus Main Theme By John Paesano

Detroit: Become Human — Kara Main Theme By Philip Sheppard

Detroit: Become Human — Connor Main Theme By Nima Fakhrara

In this episode of the Data Show, I spoke with Andrew Burt, chief privacy officer at Immuta, and Steven Touw, co-founder and CTO of Immuta. Burt recently co-authored an upcoming white paper on managing risk in machine learning models, and I wanted to sit down with them to discuss some of the proposals they put forward to organizations that are deploying machine learning.

Some high-profile examples of models gone awry have raised awareness among companies for the need for better risk management tools and processes. There is now a growing interest in ethics among data scientists, specifically in tools for monitoring bias in machine learning models. In a previous post, I listed some of the key considerations organization should keep in mind as they move models to production, but the upcoming report co-authored by Burt goes far beyond and recommends lines of defense, including a description of key roles that are needed.

Managing Risk in Machine Learning Models

Compensation is a topic near and dear to everyone’s heart… but what does “compensation” fully mean — and what does it include, what doesn’t it include? How do entrepreneurs compete for talent in an intensely competitive environment, while balancing their startup’s affordability considerations?

This wide-ranging episode of the a16z Podcast (based on an event held for entrepreneurs at Andreessen Horowitz earlier this year) covers all things compensation — from philosophical questions such as how to get to alignment around your company’s compensation philosophy to details such as the tradeoffs between RSUs vs. stock options. The discussion includes Steve Cadigan, talent advisor and cofounder at ISDI Digital University; Thanh Nguyen, Executive Director at Connery Consulting; Greg Loehmann, principal at Compensia; and a16z partner Shannon Schiltz, who heads up a16z’s human resources, tech talent, and people practices operation.

a16z Podcast: All Things Compensation

<img src="https://elroycdn.twit.tv/sites/default/files/styles/twit_album_art_144x144/public/images/shows/the_tech_guy/album_art/audio/ttg1400audio.jpg?itok=NYIIVc8B" align="right" hspace="20" vspace="20" border="0" title="The Tech Guy (Audio)" alt="The Tech Guy (Audio)"/>Using Plume for pools of WiFi in each household room, understanding Windows 10 version 1803, ensuring an updated Netgear Nighthawk AC1900 router, recommending Powerline Networking, controlling multiple TVs individually with software, the uses o af Netatmo weather station, and Chris Marquardt's appreciation for hand photos! 
Host: <a href="https://twit.tv/people/leo-laporte">Leo Laporte</a> 
Guest: <a href="https://chrismarquardt.com/" target="_blank">Chris Marquardt</a> 
Download or subscribe to this show at <a href="https://twit.tv/shows/the-tech-guy">https://twit.tv/shows/the-tech-guy</a>. 
For detailed show notes, visit <a href="http://techguylabs.com/" target="_blank">techguylabs.com</a>. 
Bandwidth for The Tech Guy podcast is provided by <a href="https://www.cachefly.com/" target="_blank">CacheFly</a>.
Sponsors:<ul>
<li><a href="http://LastPass.com/twit">LastPass.com/twit</a></li>
<li><a href="http://texture.com/twit">texture.com/twit</a></li>
</ul>

The Tech Guy 1500

In this episode of the Data Show, I spoke with Eric Jonas, a postdoc in the new Berkeley Center for Computational Imaging. Jonas is also affiliated with UC Berkeley’s RISE Lab. It was at a RISE Lab event that he first announced Pywren, a framework that lets data enthusiasts proficient with Python run existing code at massive scale on Amazon Web Services. Jonas and his collaborators are working on a related project, NumPyWren, a system for linear algebra built on a serverless architecture. Their hope is that by lowering the barrier to large-scale (scientific) computation, we will see many more experiments and research projects from communities that have been unable to easily marshal massive compute resources. We talked about Bayesian machine learning, scientific computation, reinforcement learning, and his stint as an entrepreneur in the enterprise software space.

Building Accessible Tools for Large-Scale Computation and Machine Learning

In this episode of the Data Show, I spoke with Roger Chen, co-founder and CEO of Computable Labs, a startup focused on building tools for the creation of data networks and data exchanges. Chen has also served as co-chair of O'Reilly's Artificial Intelligence Conference since its inception in 2016. This conversation took place the day after Chen and his collaborators released an interesting new white paper, "Fair value and decentralized governance of data." Current-generation AI and machine learning technologies rely on large amounts of data, and to the extent they can use their large user bases to create “data silos,” large companies in large countries (like the U.S. and China) enjoy a competitive advantage. With that said, we are awash in articles about the dangers posed by these data silos. Privacy and security, disinformation, bias, and a lack of transparency and control are just some of the issues that have plagued the perceived owners of “data monopolies.”

Acquiring and Sharing High-Quality Data

In this episode of the Data Show, I speak with Cassie Kozyrkov, technical director and chief decision scientist at Google Cloud. She describes "decision intelligence" as an interdisciplinary field concerned with all aspects of decision-making, and which combines data science with the behavioral sciences. Most recently she has been focused on developing best practices that can help practitioners make safe, effective use of AI and data. Kozyrkov uses her platform to help data scientists develop skills that will enable them to connect data and AI with their organizations' core businesses.

We had a great conversation spanning many topics, including:

How data science can be more useful

The importance of the human side of data

The leadership talent shortage in data science

Is data science a bubble?

Make Data Science More Useful

In this episode of the Data Show, I speak with Alex Ratner, project lead for Stanford’s Snorkel open source project; Ratner also recently garnered a faculty position at the University of Washington and is currently working on a company supporting and extending the Snorkel project. Snorkel is a framework for building and managing training data. Based on our survey from earlier this year, labeled data remains a key bottleneck for organizations building machine learning applications and services.

Ratner was a guest on the podcast a little over two years ago when Snorkel was a relatively new project. Since then, Snorkel has added more features, expanded into computer vision use cases, and now boasts many users, including Google, Intel, IBM, and other organizations. Along with his thesis advisor professor Chris Ré of Stanford, Ratner and his collaborators have long championed the importance of building tools aimed squarely at helping teams build and manage training data. With today’s release of Snorkel version 0.9, we are a step closer to having a framework that enables the programmatic creation of training data sets.

Labeling, Transforming, and Structuring Training Data Sets for Machine Learning

In this episode of the Data Show, I speak with Kesha Williams, technical instructor at A Cloud Guru, a training company focused on cloud computing. As a full stack web developer, Williams became intrigued by machine learning and started teaching herself the ML tools on Amazon Web Services. Fast forward to today, Williams has built some well-regarded Alexa skills, mastered ML services on AWS, and has now firmly added machine learning to her developer toolkit.

Becoming a Machine Learning Practitioner

In this episode of the Data Show, I speak with Michael Mahoney, a member of RISELab, the International Computer Science Institute, and the Department of Statistics at UC Berkeley. A physicist by training, Mahoney has been at the forefront of many important problems in large-scale data analysis. On the theoretical side, his works spans algorithmic and statistical methods for matrices, graphs, regression, optimization, and related problems. On the applications side, he has contributed to systems used for internet and social media analysis, social network analysis, as well as for a host of applications in the physical and life sciences. Most recently, he has been working on deep neural networks, specifically developing theoretical methods and practical diagnostic tools that should be helpful to practitioners who use deep learning.

Understanding Deep Neural Networks

O'Reilly Media spreads the knowledge of innovators. At O’Reilly, a big part of our business is paying attention to what’s new and interesting in the world of technology. We have a pretty good record at having anticipated some of the big technology developments in recent history. For instance, we launched the first commercial Web site, GNN, in 1993; we organized the meeting at which the term “open source” was first adopted; we were early investors in Blogger, which helped launch the blogging revolution; and more recently, our Web 2.0 conference launched a world-wide meme. We call this predictive sense the “O’Reilly Radar.” And while we’re certainly not always right, we are, at least, good at making interesting guesses.

Our methodology is simple: we draw from the wisdom of the alpha geeks in our midst, paying attention to what’s interesting to them, amplifying these weak signals, and seeing where they fit into the innovation ecology. Add to that the original research conducted by our Research team, and you start to get a good picture of what the technology world is thinking about.

Managing Risk in Machine Learning Models

O'Reilly Radar