KDnuggets™ News 12:n26, Nov 14

Features

Poll Results: Responsibility for Predictions ( comments) - Nov 13, 2012.
KDnuggets readers are almost equally split, with 45% voting that there is no responsibility for predictions, and an almost equal number saying that there should be some responsibility. Asian and East European data miners had the highest vote for responsibility, and Latin American ones - the lowest. What do you think?
Nate Silver gets Gold on correct predictions of US 2012 Elections ( comments) - Nov 7, 2012.
The contest between pundits who criticized Nate Silver (and other statisticians) has been won decisively by the statisticians.
Stanford Data Mining Certificates Online - Nov 6, 2012.
The data mining industry is moving fast. Navigate your career across any terrain with Stanford University online graduate courses and certificates in Data Mining and Applications, and Mining Massive Data Sets.
PAW: Predictive Analytics World London 2012, Nov 27-28 - Nov 13, 2012.
For those not yet convinced this two-day trip to London is worthwhile, here are Dr. Geert Verstraeten personal top three reasons to attend Predictive Analytics World London 2012
Big Data Innovation Summit, Las Vegas, January 30-31 - Nov 8, 2012.
The Summit at Caesar's Palace will bring together the forward-thinking executives leading data science initiatives for two days of insight and unparalleled networking opportunities.
Top news for Nov 4-10: Nate Silver gets Gold on election predictions; Switching Detection Challenge - Nov 11, 2012.
Nate Silver gets Gold on correct election predictions; Data crunchers who helped Obama win; Yandex Switching Detection Challenge;
Top jobs: Data Scientist at Salesforce; Postdoc, Bayesian Inference/ML at Ohio State U
Top news for Oct 28 - Nov 3: Responsibility for predictions? Strata NYC Slides/Video; GMDH Shell - Nov 4, 2012.
Data Scientists / Data Miners responsible for their predictions? Data Mining Cartoon; Strata NYC Slides/Video; GMDH Shell
Top jobs: Data Mining Engineer at Groupon; Sr Business Data Analyst at Jack in the Box
Top news, jobs in October: Data Mining Cartoon; Predicting Obama vs Romney; StatSoft free software to struggling EU countries - Nov 1, 2012.
Data Mining Cartoon: Twitter, Facebook, and Prediction; New Poll: Predicting Obama vs Romney; StatSoft offers free software to struggling European Countries
Top jobs: Data Scientist at Kyruus; Associate Predictive Modeler at Allstate;
Additions to KDnuggets Directory in October - Nov 1, 2012.
New companies, datasets, education, meetings, publications, software, solutions

Software

IBM Text Analytics Made Easy ... See How - Nov 7, 2012.
Tired of reading, recording and manually coding the text responses to your surveys? Discover how to make open-end text analysis easier and more efficient. Download the white paper now.
IBM: A Better Way to Collect the Data You Need - Oct 31, 2012.
Download this white paper to discover how IBM SPSS Data Collection can help you field surveys to get the answers you need easily and efficiently - providing you with the data you need to solve problems and make better decisions. Download now.
GMDH Shell Forecasting and Data Mining Software - Nov 1, 2012.
GMDH Shell lets you do forecasting or data mining tasks with minimum effort, thanks to the unique GMDH forecasting method and state-of-the-art parallel processing technology.
Datameer Analytics Apps Market - Nov 12, 2012.
Built by data scientists, analysts, or subject matter experts, analytic apps range from horizontal use cases like email and social sentiment analysis to vertical or even product-specific applications like advanced Salesforce.com sales-cycle analysis.
Mikut Data Mining Tools spreadsheet ( comments) - Nov 9, 2012.
Updated table describing 294 current and historical data mining tools is now online in Excel format
Predixion Software Free Predictive Analytics Developers Edition - Nov 7, 2012.
Predixion Insight provides easy to develop and deploy enterprise class predictive analytics utilizing Microsoft Excel, PowerPivot and SharePoint.
scikit-learn: machine learning in Python - Nov 5, 2012.
scikit-learn is a Python module integrating machine learning, data mining, and visualization with Big Data and scientific Python tools numpy and scipy.

Courses, Events

New Accredited Online MS in Health and Life Science Analytics - Nov 7, 2012.
This degree from NU provides students with a strong foundation in analytics through hands on use of SAS and real-world data withcourses spanning data management, predictive modeling, data mining, health management analytics, clinical and medical research, and more.
Salford Data Mining Training, Dec 5-7, San Diego: Predictive Modeler, CART, MARS, RandomForests, and more - Nov 5, 2012.
Training covers Data Mining with Decision Trees (CART), Predictive Modeling with MARS Automated Non-linear Regression and RandomForests, and Advanced Data Mining: TreeNet, GPS, Rulefit and Interaction Detection.
Wharton Course: Bringing Customer Lifetime Value to Life: Practical Methods and Applications - Nov 3, 2012.
Customer Lifetime Value (CLV) is a critical concept for virtually every organization that wants to be customer centric. This 2-day workshop helps clarify key issues and shows how to compute and utilize CLV in a practical business setting.

Webcasts

November 15 Webcast - Analytically Speaking featuring Michael Berry and Gordon Linoff - Nov 6, 2012.
With examples from their work with companies like TripAdvisor, The New York Times and T-Mobile, these experts will share insights on improving cross-sell, analyzing acquisition channels over time, scoring customers for attrition risk, and much more.

Meetings

IEEE ICDM 2012, Brussels, Dec 10-13, Early Reg by Nov 18 - Nov 9, 2012.
IEEE ICDM is a leading research conference in data mining. Highlights of the program include invited talks by Martin Krzywinski (Genome Science Center), Bernhard Schoelkopf (Max Planck Institute), and Jure Leskovec (Stanford University), three tutorials, the ICDM contest, 16 workshops, 6 demos, an Industry and Government track and a panel.

Jobs

Imaging Scientist at Aware, Bedford, MA - Nov 13, 2012.
A self-motivated, team-oriented experienced research scientist or research engineer with experience in image analysis, a solid understanding of pattern recognition concepts and strong algorithm development expertise.
Senior Predictive Modeler at Marshall & Swift/Boeckh (MSB), Milwaukee, WI - Nov 9, 2012.
Develop predictive models and advanced risk analytics for our clients in the insurance industry, and let us develop you.
Data Scientist at MaxPoint Interactive, Morrisville, NC - Nov 8, 2012.
The ideal candidate is an unabashed data geek, who enjoys searching for datasets to explore and mashup, and is equally interested in gleaning insights from the data that have business application.
Senior Applied Statistician at MaxPoint Interactive, Morrisville, NC - Nov 8, 2012.
We are not looking for someone to generate reports and PowerPoint presentations. We need smart, quantitative people to rip through data and find novel insights that will help us grow.
Lead Analytic Scientist at Verisk Health, South Jordan, UT - Nov 6, 2012.
Join a dynamic, growing company providing important tools to control fraud, waste and abuse in health care and experience the exciting growth of Utah "Silicon Slopes" along the Wasatch Range.
Customer Intelligence Manager, Data Scientist at Salesforce, San Francisco, CA - Nov 5, 2012.
Work with massive datasets from multiple sources and use statistics, data mining and machine learning techniques to develop algorithms that analyze customer usage patterns and make actionable recommendations
Research Engineer in AI, Data Mining, Knowledge Discovery at Belcurves Technologies, Nice, France - Nov 2, 2012.
Join our dynamic RnD team for the development of a new automatic classification system. Lead the design and development of modules for the knowledge extraction and the automatic classification of text and 2D shapes.
Customer Value Analyst at Mercury, Denver/Durango, CO - Nov 2, 2012.
Do you enjoy solving challenging business problems that combine quantitative analysis, scientific investigation, and real-world customer impact?
Director, Data Science at MarketShare, LA, SF or NYC - Nov 1, 2012.
Lead MarketShare research on attribution, optimization and other advertising analytics modeling problems. Be a key leader on the team that develops and optimizes actionable analytics solutions.

Academic/Research positions

Faculty, Business Analytics, Data/Text Mining, or Large Data at UIC, Chicago, IL - Nov 12, 2012.
Preference is for candidates with strong commitment to research in business analytics, data/text mining, or large data approaches for research, and interest in one or more related areas like business intelligence, social media, e-business, supply chain, risk management, health informatics.
Postdoc in Clinical NLP and Data Mining at Cincinnati Children's (CCHMC), Cincinnati, OH - Nov 11, 2012.
Do you want to apply your computational skills, machine learning and NLP experience to solve real world challenges to help patients by automating the clinical decision making?
Postdoc, Bayesian Inference/Machine Learning at Ohio State U., Columbus, OH and Boston, MA - Nov 8, 2012.
Help develop novel adaptive psychophysical tests and analyze large-scale longitudinal datasets on NEI-funded project that examine the visual deficits in adults and children with amblyopia.
Postdocs, machine learning, text mining at U. of Milano-Bicocca, Milan, Italy - Nov 5, 2012.
2 postdocs focusing on eJRM project, the development of automatic annotation and classification techniques of judicial textual documents. Apply by Nov 22.
Doctoral scholarships, Information Technology / Computer Science at DIPF, Frankfurt am Main - Nov 2, 2012.
Scholarships are granted for completing a doctoral thesis in information technology/computer science focusing on text or data mining, automatic language processing, computer science or a related topic in the field of "Knowledge Discovery in Scientific Publications". Applications due Nov 28.
Research/Lead Scientist at DIPF, Frankfurt am Main, Germany - Nov 1, 2012.
The position focus will be on the automatic text assessment and classification in innovative applications of computer science to the educational content.

Competitions

WSDM Yandex Switching Detection Challenge - Nov 5, 2012.
The focus is on predicting the search engine switching events using user logged search behavior context, using a fully anonymized dataset shared by Yandex
SBP 2013 Data Challenge: Mobile phones - Oct 31, 2012.
Use MIT Human Dynamics/Media Lab mobile phone data sets on "Reality Commons", to propose and submit applications that show scientific and business values of these data.

Publications

Book Draft: Introduction to Machine Learning - Nov 9, 2012.
Over the past two decades Machine Learning has become one of the mainstays of information technology and with that, a rather central, albeit usually hidden, part of our life.
Big Data Journal Preview - Nov 9, 2012.
Big Data Journal had a very successful launch of the journal at the Strata + Hadoop World conference with an impressive preview issue and plans are on track for an inaugural issue in early 2013.
Data Crunchers Who Helped Obama Win - Nov 8, 2012.
Time magazine looks inside the secret world of the data crunchers who helped Obama win. The magic tricks that opened wallets and how they were repurposed to turn out votes. The top performing campaigns were up to 10 times better than underperformers.
New Book: Bisociative Knowledge Discovery - Nov 7, 2012.
The focus of this book is a network-based integration of various types of data repositories and the development of new ways to analyse and explore the resulting gigantic information networks.
Tim Graettinger: Data Mining Misconceptions The 50-50 Problem ( comments) - Nov 6, 2012.
When creating a predictive model, we need to "tune" it to match our client's concern between false alarms and missed opportunities. Setting the cut-off point between 'promising' and 'unpromising' depends a lot on our client's biggest concern -- missed opportunities or false alarms.
Better Learning via Educational Data Mining and Analytics ( comments) - Nov 6, 2012.
Dept. of Education examines how Educational data mining and learning analytics can be used to improve teaching and learning
Strata and Hadoop World NYC 2012 Slides and Video ( comments) - Oct 31, 2012.
A treasure trove of presentations and selected videos from the great Strata 2012 NYC conference, covering Analytics, Big Data, Hadoop, Case Studies, Human Face of Big Data and lots more.
Top KDnuggets tweets, Nov 8-11: Making Statistics Simpler - explaining Bayes Rule; Ramp, a python library for rapid machine learning - Nov 12, 2012.
Making Statistics Simpler - explaining Bayes Rule; Ramp, a python library for rapid machine learning; Book Draft: Intro to Machine Learning; Twotorials: 2-minute tutorials on R, how to install, code, graph, and analyze
Top KDnuggets tweets, Nov 5-7: WSDM Yandex Switching Detection Challenge; Infographic: Data Science is more than mining - Nov 8, 2012.
WSDM Yandex Switching Detection Challenge; Infographic: Data Science is more than mining; Stanford Online: Data Mining Certificates - earn world-class credentials; Data Scientist positions at the CIA
Top KDnuggets tweets, Nov 1-4: Best new jobs in America: Data Scientist; MIT Researchers predict Twitter Trends - Nov 5, 2012.
Best new jobs in America: Data Scientist (#3 on CNN list); MIT Researchers predict Twitter Trends w. 95% accuracy, better than Twitter; Data Mining with R videos, from Georgia R school; Data Mining Book Review: Dance with Chance - how to avoid being fooled by chance
Top KDnuggets tweets, Oct 29-31: Open Library: free downloads of 1M eBooks; Analyzing BigData is returning an edge to Microsoft - Nov 1, 2012.
Open Library: free downloads of more than 1M eBooks; Analyzing BigData is returning an edge to Microsoft - next products to use Machine Learning; Advice for the aspiring Data Scientist; Researchers win Kaggle Molecular Activity Challenge w. deep neural networks

News Briefs

Nate Silver on Colbert Report predicts US election - Nov 7, 2012.
Election commentators botch metaphors, Math has a liberal bias, and New York Times analyst Nate Silver on predicting elections.
IBM Big Data Analytics Splash - Nov 2, 2012.
IBM Information on Demand 2012 and Business Analytics Forum showcased a broad range of big data and analytics tools and services, including Cognos Insight Personal Edition (free for personal use) and Analytics Answers from the cloud.
IEEE ICDM 2012 Research and Service Awards - Nov 1, 2012.
The IEEE ICDM Research Contributions Award recognizes influential contributions to the field of data mining, and Outstanding Service Award recognizes major service contributions data mining as a field and ICDM conference in data mining.
Aberdeen Group: Uncovering Customer Insight in Big Data - Nov 3, 2012.
Enterprise leaders in big data programs are beginning to differentiate themselves from competitors and slow adopters by matching advanced analytic plans with a blend of specialists and business demands. Leaders could meet time-to-information window 82% of the time, compared to 62% for followers.

CFP - Calls for Papers

DANTH 2013: Data Analytics for Targeted Healthcare , due Dec 30
BSI2013: Behavior and Social Informatics, due Jan 6

Quote

Nate Silver gets Gold on correct predictions of US 2012 Elections