Home » Posts tagged 'business analytics'
Tag Archives: business analytics
The R Users of New Delhi met for the second time on Dec 15, 2012. We meet on the third Saturday of every month.
We talked on epidemiology using epi calc package ( we have 1 doctor and 1 bio statistician) , and Cloud Computing ( we have two IT guys) and Business Analytics. We also discussed the GUI , R Commander , Rattle, and Deducer for beginners and people transitioning to R from other analytics software. We also discussed the R for SAS and SPSS Users books, and R for Data Mining Book. The free book for R for Epidemiology ( http://cran.r-project.org/doc/contrib/Epicalc_Book.pdf ) was mentioned . Not bad for 1 hour.
We are currently unfunded and unsponsored , I hope to get some sponsors to give away R books to encourage users and group members (excluding my own). The only catch to join this meetup group, you either need to attend (and be local) or present something ( if you are not in Delhi)
I have been trying to get this group to go from Vector to Matrix to get a bigger sponsorship from Revolution , but I am constrained by meeting in a public cafe. That is due to change since we managed to get one sponsor for meeting place in Noida ( a Business School batchmate who owns his office)
Deadlines for applications are:
- March 31, 2013 for Matrix and Array level groups.
- September 30, 2013 for Vector level groups.
2013 Sponsorship Levels
The size of the annual grant depends on the size of your group.
|Level||For groups that are:||Requirements||Annual Grant ($USD)|
|Vector||Just getting started||A group name, group webpage, and a focus on R. (Here are some tips on starting up a new R user group.)||$100|
|Matrix||Smaller but established||3 meetings in last 6 months with 30 attendees or more.||$500|
|Array||Larger and groups||3 meetings in last 6 months with 60 attendees or more.||$1000|
- New Delhi R User group meets up (decisionstats.com)
BigML has created a marketplace for selling Datasets and Models. This is a first (?) as the closest market for Predictive Analytics till now was Rapid Miner’s marketplace for extensions (at http://rapidupdate.de:8180/UpdateServer/faces/index.xhtml)
SELL YOUR DATA
You can make your Dataset public. Mind you: the Datasets we are talking about are BigML’s fancy histograms. This means that other BigML users can look at your Dataset details and create new models based on this Dataset. But they can not see individual records or columns or use it beyond the statistical summaries of the Dataset. Your Source will remain private, so there is no possibility of anyone accessing the raw data.
SELL YOUR MODEL
Now, once you have created a great model, you can share it with the rest of the world. For free or at any price you set.Predictions are paid for in BigML Prediction Credits. The minimum price is ‘Free’ and the maximum price indicated is 100 credits.
White Box Models
Clicking on the white open lock will open up your model to the rest of the world. Anyone can now buy your model, explore it, use it to make predictions
Black Box Models
If you choose the black box setting (the black open lock icon), other BigML users will NOT be able to view or clone your model, but they will be able to use it to make predictions.
DOWNLOAD YOUR MODEL
BigML.com have added downloads to our models. Simply choose the format you want and you can copy/paste the code or text. There is a range of formats that they offer currently: JSON PML, PMML, Python, Ruby, Objective-C, Java, the rules of the decision tree in plain text and a Summary overview of your model. Around the corner are MS Excel downloads and R (of course!).
PUBLICIZE YOUR MODEL
There’s also an ‘embed’ function, so now you can embed the little poster of your model in your blog post or website, so it is easy to share it in your own environment.
It is nice to see Models and Data getting the APPY treatment and hopefully, it will encourage other vendors Iike Google Prediction API etc to further spend thought and effort to reward data mining individuals directly without going through corporate intermediaries while ensuring intellectual property safeguards .
An R package market for enterprises? for Python libraries? JMP addins? A market for SAS Macros- who knows what the future shall hold. But overall, this is a very positive step by the BigML.com team. The App marketplace has helped revolutionize mobile and desktop computing and hopefully it will do the same for Business Analytics.
Analytics 2012 Conference
SAS and more than 1,000 analytics experts gather at
Analytics 2012 Conference Details
Pre-Conference Workshops – Oct 7
Conference – Oct 8-9
Post-Conference Training – Oct 10-12
Caesars Palace, Las Vegas
The following are confirmed keynote speakers for Analytics 2012. Since he co-founded SAS in 1976, Jim Goodnight has served as the company’s Chief Executive Officer.
Dr. William Hakes is the CEO and co-founder of Link Analytics, an analytical technology company focused on mobile, energy and government verticals.
Tim Rey has written over 100 internal papers, published 21 external papers, and delivered numerous keynote presentations and technical talks at various quantitative methods forums. Recently he has co-chaired both forecasting and data mining conferences. He is currently in the process of co-writing a book, Applied Data Mining for Forecasting.
Plan to come to Analytics 2012 a day early and participate in one of the pre-conference workshops or take a SAS Certification exam. Prices for all of the preconference workshops, except for SAS Sentiment Analysis Studio: Introduction to Building Models and the Business Analytics Consulting Workshops, are included in the conference package pricing. You will be prompted to select your pre-conference training options when you register.
Sunday Morning Workshop
SAS Sentiment Analysis Studio: Introduction to Building Models
This course provides an introduction to SAS Sentiment Analysis Studio. It is designed for system designers, developers, analytical consultants and managers who want to understand techniques and approaches for identifying sentiment in textual documents.
Sunday, Oct. 7, 8:30a.m.-12p.m. – $250
Sunday Afternoon Workshops
Business Analytics Consulting Workshops
This workshop is designed for the analyst, statistician, or executive who wants to discuss best-practice approaches to solving specific business problems, in the context of analytics. The two-hour workshop will be customized to discuss your specific analytical needs and will be designed as a one-on-one session for you, including up to five individuals within your company sharing your analytical goal. This workshop is specifically geared for an expert tasked with solving a critical business problem who needs consultation for developing the analytical approach required. The workshop can be customized to meet your needs, from a deep-dive into modeling methods to a strategic plan for analytic initiatives. In addition to the two hours at the conference location, this workshop includes some advanced consulting time over the phone, making it a valuable investment at a bargain price.
Sunday, Oct. 7; 1-3 p.m. or 3:30-5:30 p.m. – $200
Demand-Driven Forecasting: Sensing Demand Signals, Shaping and Predicting Demand
This half-day lecture teaches students how to integrate demand-driven forecasting into the consensus forecasting process and how to make the current demand forecasting process more demand-driven.
Sunday, Oct. 7; 1-5 p.m.
Forecast Value Added Analysis
Forecast Value Added (FVA) is the change in a forecasting performance metric (such as MAPE or bias) that can be attributed to a particular step or participant in the forecasting process. FVA analysis is used to identify those process activities that are failing to make the forecast any better (or might even be making it worse). This course provides step-by-step guidelines for conducting FVA analysis – to identify and eliminate the waste, inefficiency, and worst practices from your forecasting process. The result can be better forecasts, with fewer resources and less management time spent on forecasting.
Sunday, Oct. 7; 1-5 p.m.
SAS Enterprise Content Categorization: An Introduction
This course gives an introduction to methods of unstructured data analysis, document classification and document content identification. The course also uses examples as the basis for constructing parse expressions and resulting entities.
Sunday, Oct. 7; 1-5 p.m.
Introduction to Data Mining and SAS Enterprise Miner
This course serves as an introduction to data mining and SAS Enterprise Miner for Desktop software. It is designed for data analysts and qualitative experts as well as those with less of a technical background who want a general understanding of data mining.
Sunday, Oct. 7, 1-5 p.m.
Modeling Trend, Cycles, and Seasonality in Time Series Data Using PROC UCM
This half-day lecture teaches students how to model, interpret, and predict time series data using UCMs. The UCM procedure analyzes and forecasts equally spaced univariate time series data using the unobserved components models (UCM). This course is designed for business analysts who want to analyze time series data to uncover patterns such as trend, seasonal effects, and cycles using the latest techniques.
Sunday, Oct. 7, 1-5 p.m.
SAS Rapid Predictive Modeler
This seminar will provide a brief introduction to the use of SAS Enterprise Guide for graphical and data analysis. However, the focus will be on using SAS Enterprise Guide and SAS Enterprise Miner along with the Rapid Predictive Modeling component to build predictive models. Predictive modeling will be introduced using the SEMMA process developed with the introduction of SAS Enterprise Miner. Several examples will be used to illustrate the use of the Rapid Predictive Modeling component, and interpretations of the model results will be provided.
Sunday, Oct. 7, 1-5 p.m
.Of course My usual suspects for Time Series Readings are -
1) The seminal pdf (2008!!) by a certain Prof Hyndman
2) JSS Paper -Automatic Time Series Forecasting: The forecast
Package for R http://www.jstatsoft.org/v27/i03/paper
3) The CRAN View http://cran.r-project.org/web/views/TimeSeries.html
This is cluttered and getting more and more cluttered. Some help on helping recent converts to R, especially in the field of corporate forecasting or time series for business analytics would really help.
Avril does an awesome job with this curiously named ( ) booklet at http://a-little-book-of-r-for-time-series.readthedocs.org/en/latest/src/timeseries.html
Increasingly Big Data is used in writing where Business Analytics was used, and data mining is thrown in as a word just to keep liberal art majors happy that they are reading a scientific article.
Some Big Words I have noticed in my Short life-
Big Data? High Performance Analytics? High Performance Computing ? Cloud Computing? Time Sharing? Data Mining? SEMMA? CRISP-DM? KDD? Business Intelligence? Business Analytics and Optimization? (pick a card and any card)
(or Just Moore’s Law catching up with the analytics)
Replace Big Data with Analytics in these articles and let me know if you can make out much of a difference
- Big Data on Campus
- From the man who famously said BI is dead, is now burying Business Analytics within the new buzzword , SAS CMO Jim Davis
How to transform big data from an obstacle into an asset
(Related- Is big data over hyped? by Jim Davis
I am sure by 2015, Jim Davis, NYT and the merry men of analytics will find some other buzzwords to rally the troops. In the meantime, let me throw out the flag and call it Big .
While SAS language has a beautifully designed ODS (Output Delivery System) for saving output from certain analysis in excel files (and html and others), in R one can simply use the object, put it in a write.table and save it a csv file using the file parameter within write.table.
As a business analytics consultant, the output from a Proc Means, Proc Freq (SAS) or a summary/describe/table command (in R) is to be presented as a final report. Copying and pasting is not feasible especially for large amounts of text, or remote computers.
Using the following we can simple save the output in R
#We shifted the directory, so we can save output without putting the entire path again and again for each step.
#I have found the summary command most useful for initial analysis and final display (particularly during the data munging step)
# I assigned a new object to the analysis step (summary), it could also be summary,names, describe (HMisc) or table (for frequency analysis),
Note: This is for basic beginners in R using it for business analytics dealing with large number of variables.
If you have a large number of files in a local directory to be read in R, you can avoid typing the entire path again and again by modifying the file parameter in the read.table and changing the working directory to that folder
and so on…
maybe there is a better approach somewhere on Stack Overflow or R help, but this will work just as well.
you can then merge the objects created ajayt1 and ajayt2… (to be continued)