Two trends have been rapidly bringing the cost of machine learning down. For one, computing power is getting cheaper, as it always does. For the second, machine learning algorithms are becoming productized.
What this means is that many smaller scale business functions are about to feel the effects of machine learning. When it costs a million dollars to build an algorithm, only the largest companies apply machine learning to classifying their support tickets, organizing their sales database, or handling collections. But when it costs twenty dollars a month, everyone will do it.
Design Pattern 1: Training Data > Supervised machine learning requires "training data" for every individual task their told to perform and it needs to be updated each time a new initiative is integrated. And it is humans that can better refine the training data. Design Pattern 2: Human-in-the-loop > Instead of machine learning replacing one job function at a time, machine learning actually replaces a job function. This makes the person doing the job increasingly more efficient. In some cases, this can lead to fewer jobs, but in others, this can create new markets and create more jobs for the same type of work. Design Pattern 3: Active Learning > Active learning is a design pattern that combines the first two patterns. The training data collected by the Human in the Loop can be fed into the algorithm to make it better. Algorithms learn like people ; novel, complicated situations help them learn much faster.
[author] analsyzed ata from 26,000 professional matches from 2009 to 2015. Which is really a lot. Bookmakers adjust the odds depending on how people are betting on a match. So [he] looked at the odds that 7 major sports books initially offered for each match, and then compared them with the final odds, to see how far they had shifted.
Via Karen Bastien
Well?.. he might have found at least one :)
"For one player, I identified 16 matches for which bookmakers revised his odds of winning downward by at least 10 percentage points.
He lost 15 of the 16 matches, including some in which he started as a heavy favorite."
David M. Hillis, Derrick Zwickl, and Robin Gutell, University of Texas. About this Tree: This tree is from an analysis of small subunit rRNA sequences sampled from about 3,000 species from throughout the Tree of Life. The species were chosen based on their availability, but we attempted to include most of the major groups, sampled very roughly in proportion to the number of known species in each group (although many groups remain over- or under-represented). The number of species represented is approximately the square-root of the number of species thought to exist on Earth (i.e., three thousand out of an estimated nine million species), or about 0.18% of the 1.7 million species that have been formally described and named.
I don't know if the "groups" they started with are the most relevant, but I guess they did their homework and I find this visualization very interesting as it highlights how marginal human beings are on this planet.
Twitch has grown so quickly this year that it’s hard to keep track of all the amazing subgroups and communities that call Twitch home. To illustrate this, our Science team has recently been building visual maps of the Twitch world and we’re thrilled to share them with you! NOTE: Visualization and layout were completed using an open source tool called Gephi. Click any image to enlarge.
Tableau [is] a software company that turns heaps of data into visualizations for the common man: teachers, doctors, journalists, you name it. To make those tools clearer and cleaner, they recently partnered with Stamen Design, to release three new map templates, which anyone can play around with by downloading Tableau’s free software.
"the entire point of Tableau’s software update is to create a sophisticated cartography tool for anyone. Map-makers simply upload a .csv file of data into Tableau, and then get to select a template from Stamen’s canon of maps."
Research into connections between gun violence and video games to violence in real life has been inconclusive. As the hype for the latest in the Grand Theft Auto series builds up, a Guardian analysis of the top 50 video games sold in 2012 found more than half contain violent content labels, as assigned by an independent video game rating board (ESRB). Overall, one-third have weapons that depict real-life firearms
interesting visualization. Interactive roll-over feature improve readability a lot!
As for the topic, it obviously demonstrates what Miyamoto recently said about the video game industry's creativity being "immature". Indeed it is. Relying almost solely on the urge to headshot one another in various settings and with various calibers.
The Music Timeline shows genres of music waxing and waning, based on how many Google Play Music users have an artist or album in their music library, and other data (such as album release dates). Each stripe on the graph represents a genre; the thickness of the stripe tells you roughly the popularity of music released in a given year in that genre. (For example, the "jazz" stripe is thick in the 1950s since many users' libraries contain jazz albums released in the '50s.) Click on the stripes to zoom into more specialized genres.
the methodology only involve Google Play Music, but at a glance, it looks pretty accurate.
A former Frog designer visualizes the literature on the importance of play for fostering creativity in children.
" It used to be that play was just play. There wasn’t a whole lot to say about it. Kids climbed trees, hit balls, and did experiments on insects. But by the mid-20th century, children’s play was being extensively studied, classified, and taxonomied by pioneering psychologists such as Melanie Klein and D.W. Winnicott. Now the literature is so vast and complex that parents might have a hard time figuring out what's right and what's not. Laura Richardson, who spent 10 years at Frog Design, has boiled it all down into one playful infographic: The Periodic Table of 21st Century Play."
Santiago Ortiz is an infoviz researcher and inventor. He uses his background in mathematics and complexity sciences to push the boundaries of information visualization and data based story telling. He was born in Colombia, where he began creating images out of code when he was 8. In 2005 he co-founded Bestiario (Barcelona), the first company in Europe devoted to information visualization.
great presentation. His work is amazing. I'd like to see Aaron Koblin and him have a chat :)
Something about these visualizations by media artist Tobias Gremmler is making me see the martial arts anew. As spotted by Prosthetic Knowledge, motion captured from a kung fu fighter is filtered through various schemes of computer-generated graphics to produce living sculptures that are strongly reminiscent of work by Universal Everything. And as a viewer, you’re left with just the forms and none of the violence—martial art becomes abstract art.
Variation 4 .1 made me think of Aaron Koblin's excellent work on Radiohead's House of Cards music video
The business of heroes needs saving from the crushing weight of its own data. The Marvel team thinks they’ve built a solution: a massive database that uses graph theory to give fans a simple take on characters that span comics, movies, and video games.
Algorithms are a fascinating use case for visualization. To visualize an algorithm, we don’t merely fit data to a chart; there is no primary dataset. Instead there are logical rules that describe behavior. This may be why algorithm visualizations are so unusual, as designers experiment with novel forms to better communicate. This is reason enough to study them.
Mike Bostock is the author of the d3.js library ( http://d3js.org/ ). He works for the New York Times, and is at the origin of all the impressive data visualizations they do.
Here he goes at length explaining all the reasoning behing algorithms visualization. Deep. Beautiful.
Recently we released a global heatmap of 77,688,848 rides and 19,660,163 runs from the Strava dataset. This was more of an engineering challenge to create a visualization of that size than anything else. But still, the map has raised many questions about how and where people run and ride.Some of these can only be answered using the raw data, which is addressed by our Metro product.
Great heatmap visualization of where people run and bike on the planet
Each stop on the subway is a node (a place where data is sent and received like an Internet service provider) assigned to a country. Where there were multiple nodes in one country, Graham and De Sabbata combined them into one stop. The two took node data from cablemap.info. "The map ... aims to provide a global overview of the network, and a general sense of how information traverses our planet," they wrote.
Sharing your scoops to your social media accounts is a must to distribute your curated content. Not only will it drive traffic and leads through your content, but it will help show your expertise with your followers.
How to integrate my topics' content to my website?
Integrating your curated content to your website or blog will allow you to increase your website visitors’ engagement, boost SEO and acquire new visitors. By redirecting your social media traffic to your website, Scoop.it will also help you generate more qualified traffic and leads from your curation work.
Distributing your curated content through a newsletter is a great way to nurture and engage your email subscribers will developing your traffic and visibility.
Creating engaging newsletters with your curated content is really easy.