Predictive analytics and other categories of advanced analytics are becoming a major factor in the analytics market. We evaluate the leading providers of advanced analytics platforms that are used to build solutions from scratch.
Announcing the new Professional Master’s program in Big Data
The School of Computing Science at Simon Fraser University is offering a NEW Professional Masters program in Big Data. This four-semester, hands-on program will prepare you for an exciting and well-paid career as a data scientist.
This program is intended for students with some previous programming experience who wish to learn about the state-of-the-art in big data analysis.
Applications are NOW OPEN
Application deadline is April 1, 2014 for the Fall 2014 Cohort. We will keep admissions open until we fill available slots for the Fall 2014 cohort. Our goal is to notify students of acceptance to the Program by May 1st, 2014.
This interactive illustration represents how twitter’s employees interact among themselves. This stunning design was created by Santiago Ortiz. It shows vividly the pivotal employees in Twitter’s twittersphere. To view the interactive visualization, click here.
Have you ever wondered what a Twitter conversation looks like from 10,000 feet? A new report from the Pew Research Center, in association with the Social Media Research Foundation, provides an aerial view of the social media network. By analyzing many thousands of Twitter conversations, we identified six different conversational archetypes. Our infographic describes each type of conversation network and an explanation of how it is shaped by the topic being discussed and the people driving the conversation.
Conversations on Twitter create networks with identifiable contours as people reply to and mention one another in their tweets. These conversational structures differ, depending on the subject and the people driving the conversation. Six structures are regularly observed: divided, unified, fragmented, clustered, and inward and outward hub and spoke structures. These are created as individuals choose whom to reply to or mention in their Twitter messages and the structures tell a story about the nature of the conversation.
The graph represents a network of 176 Twitter users whose recent tweets contained “sunbelt14 OR sunbelt2014″, or who were replied to or mentioned in those tweets, taken from a data set limited to a maximum of 18,000 tweets. The network was obtained from Twitter on Sunday, 23 February 2014 at 16:55 UTC.
The tweets in the network were tweeted over the 7-day, 4-hour, 11-minute period from Sunday, 16 February 2014 at 12:38 UTC to Sunday, 23 February 2014 at 16:50 UTC.
There is an edge for each “replies-to” relationship in a tweet. There is an edge for each “mentions” relationship in a tweet. There is a self-loop edge for each tweet that is not a “replies-to” or “mentions”.
The graph is directed.
The graph’s vertices were grouped by cluster using the Clauset-Newman-Moore cluster algorithm.
The graph was laid out using the Harel-Koren Fast Multiscale layout algorithm.
The edge colors are based on edge weight values. The edge widths are based on edge weight values. The edge opacities are based on edge weig
The CFTC just released weekly Commitments of Traders data, which break down positioning in the futures markets for various asset classes as of February 18.
The chart shows how hedge funds are positioned in these various markets. For each asset class, it measures how many standard deviations away hedge funds are from the average position size over the last two years.
Speculative positioning in coffee futures is 4.31 standard deviations above the two-year average, whereas speculative positioning in E-mini Dow Jones Industrial Average futures is 1.99 standard deviations below the two-year average.
By this measure, hedge funds are also particularly long cattle, crude oil, milk, and natural gas.
On the flip side, they are also particularly short eurodollars, the Mexican peso, the Canadian dollar, and E-mini S&P 500 futures
U.S. top sending country; India top receiving country
The map shows population density; the brightest points are the highest densities. Each country is colored according to its average annual gross national income per capita, using categories established by the World Bank (see key below). Some nations— like economic powerhouses China and India—have an especially wide range of incomes. But as the two most populous countries, both are lower middle class when income is averaged per capita.
- Machine Learning Tools
BigML is a cloud-based machine learning platform that allows users to create visual predictive models using raw data and structured datasets. Last month, BigML announced the availability of the 2014 winter release, which includes features that boost predictive modeling. The company also introduced a new paradigm called Programmatic Machine Learning that is the “ability to programmatically transform a dataset via a high-level language and a cloud-based API together.”
The BigML API makes it possible for developers to build applications that incorporate predictive models and near real-time predictions.
Datumbox is a machine learning platform that focuses on natural language processing (NLP). The Datumbox platform features a variety of functions including sentiment analysis, Twitter sentiment analysis, language detection, educational detection and keyword extraction.
Diffbot uses computer vision, machine learning and other technologies to extract text, images, links, HTML attributes and other elements from Web pages. In August 2013, the company released the Diffbot Product API, which can extract product information from the pages of e-commerce websites. Earlier this month, ProgrammableWeb reported on the release of 35+ new Diffbot client libraries in a variety of programming languages.
The company provides a suite of Diffbot APIs for extracting data from Web page news articles, Web site home pages, e-commerce product pages and other types of Web pages. There are also APIs for extracting Web page images and automatically classifying Web page links.
Ersatz Labs is a startup and developer of a new platform called Ersatz, described by the company as “the first cloud-based neural network platform.” The Ersatz platform allows developers to build applications that utilize deep neural networks without the need to have extensive knowledge in machine learning.
There is an API that can be accessed via HTTP, and a client library in Python is also available, so Ersatz can be easily integrated with Web, mobile and desktop applications. Ersatz is currently in private beta, and developers interested in participating can request an invitation on the official company Web site.
Google Prediction API
The Google Prediction API provides developers access to Google’s cloud-based machine learning platform and pattern-matching functions. The API is used in conjunction with the Google Cloud Storage API and allows developers to incorporate functions into their apps such as sentiment analysis, spam detection, message routing decisions, suspicious activity identification and more.
IBM Watson is a machine learning platform that focuses on NLP, hypothesis generation and evidence-based learning. In November 2013, ProgrammableWeb reported that IBM had launched the Watson Developer Cloud, a cloud-based marketplace that provides access to APIs, documentation, self-service training materials and other tools for developers to build IBM Watson-powered applications.
Last month, IBM announced that the company will invest more than $1 billion in the new Watson Group, which will be based in New York City’s “Silicon Alley.” The new group will focus on developing and promoting the IBM Watson platform and cognitive technologies. IBM also announced new Watson cognitive intelligence-based services, including IBM Watson Discovery Advisor, IBM Watson Analytics and IBM Watson Explorer.
Logical Glue is a machine learning as-a-service (MLaaS) platform that features predictive model building, predictive model real-time deployment, and real-time predictive analytics. The platform is designed to predict customer behavior for many types of markets, particularly financial lending, insurance and marketing.
The Logical Glue platform is currently in private beta; however, companies can apply to participate in the beta program, which allows them access to the platform prerelease. The next release of the platform will include the Logical Glue prediction API.
Parse.ly is a predictive content optimization and analytics platform designed for blogs, news sites and other online publishers. The home page of the Parse.ly website describes the company as “The Content Performance Authority” and the platform provides users a real-time view of article traffic based on individual posts, authors, sections and referrers. The Parse.ly platform also provides views of content metrics, social network shares, site activity and other analytics.
The Parse.ly API allows developers to programmatically access platform features such as analytics, shares, referrers, real-time, search and recommendations. There are also mobile SDKs available that can be integrated into third-party apps so reader activity can be tracked.
PredictionIO is a machine learning server that allows developers to add predictive features to software, web and mobile applications. PredictionIO is open source and can be installed on a stand-alone server. There is also a cloud version available on Amazon EC2/Amazon EBS.
The PredictionIO API enables applications to collect and manage app data and add predictive features such as predict user preferences, personalized content, content discovery, content recommendations and more. ProgrammableWeb recently published an interview with Simon Chan (cofounder and CEO of PredictionIO), which covers PredictionIO features, compares other machine learning APIs and more.
SwiftKey is a developer of touchscreen keyboard applications and word prediction technology. SwiftKey’s products Keyboard, Flow and Note incorporate machine learning and SwiftKey’s language technology, available to developers via API and SDK.
A recent TechCrunch article featured SwiftKey’s word prediction technology. Nathan Matias, a PhD student at the MIT Media Lab, used SwiftKey technology to create a sonnet essentially co-authored by Shakespeare and generated entirely from the SwiftKey next word suggestions.
Which designs have most dramatically reshaped the consumer landscape since the early ’80s?
If you had to pick just five events or designs since the early ‘80s that completely reshaped the consumer landscape, what would they be? We posed this question to Ziba’s most seasoned designers and researchers, and sorted the answers into a timeline of transformation—the innovations that took us from the era of boomboxes, floppy discs and Madonna to our present, hyper-connected reality. Then we mapped out their common themes, to answer a second, more valuable question: What statements do consumers perceive as true today that would’ve been unimaginable 29 years ago