SAS Support Communities

The International Institute of Forecasters and SAS® announce two $10,000 grants to support research on forecasting. Per the announcement:   Forecasting research has seen major changes in the theoretical ideas underpinning forecasting effectiveness over the last 30 years. However, there has been less impact on forecasting practice. We aim to put this right. For the sixteenth year, the IIF, in collaboration with SAS, is proud to announce financial support for research on how to improve forecasting methods, and business forecasting practice, including organisational aspects of management of the forecasting process. This year’s support will consist of a total of two $10,000 grants in methodology and practice/management categories.   Applications should be submitted to the IIF Office by September 30, 2019. The application must include:   • Description of the project (at most 4 pages) • C.V./resume (brief, 4 page max) • Budget and work-plan for the project (brief, 1 page max)   Criteria for the award of the grant will include likely impact on forecasting methods and business applications.   Consideration will be given to new researchers, and whether supplementary funding is likely to be gained. It is also expected that the research supported by the SAS/IIF grant be presented in an International Symposium on Forecasting (ISF) organized by the IIF. The applications will be assessed through a committee appointed by the IIF directors. The results of the evaluation will be announced to the applicants within 12 weeks of the closing date.   Grant recipients are also required to author a paper reporting on their research for possible publication in the International Journal of Forecasting (IJF). Therefore it is useful to keep in mind the IJF scope.   In addition to the advances made at our academic research institutions, there is also considerable innovation coming from industry practitioners, who are encouraged to submit proposals.   Find more details about the application process, and learn about previous recipients, on the IIF website. ... View more

The Winter 2017-2018 Journal of Business Forecasting is a special issue on the future of demand planning & forecasting. It includes two articles by SAS authors:   Charlie Chase, "Real-Time Demand Execution Anticipating Demand at the Edge"   Executive Summary: Using real-time information as it is streaming in from connected devices on the Internet of Things (IoT), edge analytics is gaining attention as the IoT has become more widespread, streaming data from manufacturing machines, online purchases, and mobile and other remote devices. Using analytic algorithms as data are generated, at the edge of the corporate network, companies can set constraints to determine what information is worth sending to the Cloud, to a demand signal repository, or other data repositories for later use. The key benefit of edge analytics is the ability to analyze data as they are generated, which decreases latency in the decision-making process as the data are collected. Rather than designing consolidated data systems where all the data are sent back to an enterprise data warehouse (or data lake) in a raw state, where they have to be cleaned and analyzed before being of any value, why not do everything at the edge of the system, including demand forecasting, using advanced algorithms or machine learning? Are you stuck in a vicious cycle of planning demand using two-to-four-week-old data, or are you conducting real-time demand execution anticipating demand at the edge?   Mike Gilliland, "The Move to Defensive Business Forecasting"     Executive Summary: Despite continuing technological advances that take us to the limits of achievable accuracy, most companies still struggle with forecasting, with many forecasts less accurate than a naive model. More complex statistical modeling, by itself, does not provide the answer. Instead, significant performance improvement can come from a "defensive" approach to business forecasting, using tools like FVA analysis. With a defensive approach, organizations can identify the waste and bad practices that degrade forecasting performance, and can thereby achieve the full potential of the technological advances.   The full articles are available to JBF subscribers and Institute of Business Forecasting members.         ... View more

For 40 years SAS has delivered the cutting edge of econometrics and time series analysis through SAS/ETS, which now has over 8000 licenses worldwide.   In March 2017, we expanded our footprint with the release of SAS Econometrics 8.1, taking advantage of Viya, leveraging the speed, scalability, and elasticity of the SAS in-memory environment. This initial release included five procedures:   CCOPULA simulates multivariate distributions for copula models. CNTSELECT does regression models for integer valued dependent variables. CPANEL fits linear regression models for panel data. CQLIM does regression models for qualitative and limited dependent variables. SEVSELECT fits distributions to the size or severity of losses or other events.   Now, SAS Econometrics 8.2 adds enhancements to the original procedures, along with four new procedures:   CCDM estimates a compound distribution model. CSPATIALREG estimates linear spatial econometric models for cross-sectional data whose observations are spatially referenced. HMM estimates hidden Markov models (HMMs) where the observables are independently distributed, conditional on the hidden states that follow a Markov chain. TSMODEL provides a new environment for time series modeling and cloud computing.   Also, SAS Econometrics now lets you utilize SAS/ETS procedures through SAS Studio.   SAS Econometrics equips our customers to address difficult, real-life questions by providing techniques to model complex business and economic scenarios and analyze the dynamic impact that specific events might have over time. It can help you understand the impact that factors such as economic and market conditions, customer demographics, pricing decisions and marketing activity have on your business, providing a scientific basis for better decision making.   SAS Econometrics is designed for all analytic audiences, including economists, financial markets analysts, data scientists, forecasters, and business analysts – anyone who deals with business, economic or time series data.   For more information, see the SAS Econometrics product page and documentation. ... View more

In March 2017 we released SAS Visual Forecasting 8.1, built for Viya, and introduced an open platform – a forecasting ecosystem. SAS Visual Forecasting was designed to be highly scalable, making one pass through the data to do multiple analyses.   Then in December, we upped the game with the release of SAS Visual Forecasting 8.2.   The biggest news of this new release is that SAS Visual Forecasting now has a user interface.  The interface facilitates a process workflow approach where you can organize and structure your forecasting process.   Within your workflow you can explore and prepare data, and utilize SAS supplied or custom forecast models that you have created. The interface allows you to compare and visualize forecast results, and make overrides to the statistical forecasts.   Also new, SAS Visual Forecasting 8.2 adds an extremely flexible capability for making manual overrides to the statistical forecast generated by your models. Instead of being limited to overrides within the forecasting project hierarchy, you can now make overrides based on attributes – such as a product’s color, or customer sentiment from an analysis of online reviews.   At the core of SAS Visual Forecasting is the TSMODEL procedure. TSMODEL innovatively requires just one read of the data to do multiple analyses. Its scripting language is optimized and compiled based on the machine it is running on, and the scripts are distributed to the available hardware.   In 8.2, two packages are added to the TSMODEL procedure. The SSA package supports Singular Spectrum Analysis, a powerful tool for detecting patterns in long time series with few model assumptions. Spectral groupings can then be individually analyzed using time series techniques. The MOTIF package supports Motif Analysis, for the discovery and detection of motifs in time series data. The new TSINFO procedure handles date variables in the input data, and can automatically determine the suitability of time ID variables in an input data set.   And finally, SAS Visual Forecasting 8.2 lets you run both SAS/ETS procedures and the SAS Forecast Server Procedures (in addition to its Viya procedures) through SAS Studio.   For more information, see the SAS Visual Forecasting product page and documentation. ... View more

Forecasters and econometricians need access to data, and SAS/ETS has long provided easy access to many commercial and government databases, including:   Commercial database vendors: FAME, DRI, Standard & Poor’s (COMPUSTAT), FactSet, Haver Analytics DLX and CRSP. Federal Reserve Economic Data (FRED). US government data: Bureau of Economic Analysis, Bureau of Labor Statistics. International agency data: International Monetary Fund (IMF), Organization for Economic Cooperation and Development (OECD). SAS/ACCESS® interfaces and SAS Data Surveyors (licensed separately) provide seamless read, write and update access to other data sources.     Since the latest 2016 release, SAS/ETS now includes access to two sources of weather-related data:   World Weather Online. NOAA Severe Weather Data Inventory (SWDI) web service.   The SAS/ETS data access interface LIBNAME engines (SASERAIN and SASENOAA) provide seamless, efficient access to weather events and weather time series data supplied by World Weather Online and the NOAA Severe Weather Data Inventory web services.   The SASERAIN interface LIBNAME engine includes the following features:   enables SAS users to retrieve weather data from the World Weather Online website uses the LIBNAME statement to enable you to download World Weather Online data and to specify which weather data time series you want to retrieve based on up to nine locations works with the SAS DATA step to write the selected weather data to a SAS data set selects past weather data based on a date range within 60 days prior to today for free nonpremium data (for premium data, the range must start no earlier than July 1, 2008) selects local forecast data based on a range defined by number of days (starts today), premium returns up to 15 days, and nonpremium (free) returns up to 5 days enables you to select the frequency of data, whether daily, hourly, every three hours, or otherwise maintains the sort order, so the locations (q-codes) are sorted in the resulting SAS data set by the order specified in the QUERY= option, by date (time ID), and by variable (time series item name) works with the SAS DATA step to perform further subsetting and to store weather data in a SAS data set supports TLS connectivity by obtaining a secure connection using the CONNECT method (if necessary) and a PROXY creates an XML map of the data for dynamic, flexible association of SAS formats and informats for all variables   The SASENOAA interface LIBNAME engine includes the following features:   enables SAS users to access severe weather data sets, such as those for tornado vortex signatures (NX3TVS), storm cell structure (NX3STRUCTURE), and preliminary local storm reports (PLSR) works with the SAS DATA step to write the selected NOAA data to a SAS data set selects data based on geospatial limits, such as by a bounding box or a centerpoint-radius combination selects data based on a date range returns data in these formats: XML; data are returned in XML format KMZ; data are returned in zipped KML format for Google My Maps (plot data on a map) SHP; mapping data are returned in zipped Esri format (four files returned inside ZIP file) works with the SAS DATA step to perform further subsetting and to store the resulting time series in a SAS data set supports TLS connectivity by obtaining a secure connection using the CONNECT method (if necessary) and a PROXY creates an XML map of the data for dynamic, flexible association of SAS formats and informats for all variables ... View more

To properly evaluate (and improve) forecasting performance, we recommend our customers use a methodology called Forecast Value Added (FVA) analysis. FVA lets you identify forecasting process waste (activities that are failing to improve the forecast, or are even making it worse). The objective is to help the organization generate forecasts that are as accurate as can reasonably be expected (given the nature of what they are forecasting), and do this as efficiently as possible (using the fewest resources).   In its simplest form, FVA compares the accuracy of a Naive forecast to the organization’s current forecasting method (usually some form of statistical forecasting with manual adjustments, or even an entirely manual process). FVA can also compare performance to alternative methods, like an automated statistical forecast (which can be generated from software such as SAS® Forecast Server or SAS® Forecasting for Desktop).   Naïve forecast: The “no change” model is standard, where the forecast is that there will be no change from the latest observation. If the organization has a supply lead time, such as two months, then the “no change” forecast for a particular month will be the actual observed value from two months prior.  So if 100 units were sold in November, the forecast for January is 100. If 125 are sold in December, then the February forecast is 125, and so on. For highly seasonal data, you may instead use a seasonal "no change" from the corresponding period the prior year. So the forecast for week 5 of 2017 is the actual from week 5 of 2016. Manual forecast: Some organizations do not use forecasting software, and use an entirely manual process where the forecast is based on management judgment. Statistical forecast: Generated by statistical models built by forecasters in forecasting software, based on historical sales alone, or by including additional variables (like pricing, promotions, events, etc.). Automated forecast: Generated entirely automatically by forecasting software with no human intervention (no tuning the models, and no manually adjusting the forecasts).   In the latter two cases, there is often the option to make manual adjustments to the computer generated forecast.   The accuracy of the Naïve forecast serves as a basis for comparison against all other forecasting activities. The Naïve forecast can be created automatically at virtually no cost, so it is important for the organization to understand what accuracy the Naïve forecast can be expected to achieve. If the Naïve’s accuracy is “good enough” for the organization’s planning and decision making purposes, then it makes sense to just use it, and stop doing any manual or statistical forecasting efforts. (Why spend time and money on a costly forecasting process if the Naïve can generate “good enough” forecasts for free?)   In most situations, however, the organization seeks forecasts that are more accurate than what the Naïve can achieve. But a typical forecasting process, even when aided by statistical forecasting software, can consume a lot of management time, at considerable cost.   The FVA approach lets you focus your improvement efforts in areas most in need of improvement – such as those where your accuracy is worse than the Naïve forecast.   Many companies find that overall, their forecasting process is performing better than just using the Naïve forecast. However, there are usually specific areas (products, locations, or other points in the organization’s forecasting hierarchy) where the Naïve forecast performs better. These should be investigated, to see if there are explainable reasons why the forecast is worse than the Naïve, and whether it can be improved. (Often the cause of such poor forecasting is political pressures within the organization. The forecast represents what management wants to happen, rather than being an unbiased best guess of what really will happen.)   Upon doing FVA analysis, a surprisingly large number of companies find that overall they are forecasting worse than doing nothing and just using the Naïve forecast! If the forecasting process cannot be improved in these areas, then it is simply wasted effort, and should be eliminated in favor of using the Naïve.   Organizations often find that, overall, Automated forecasts are more accurate than what their current forecasting process produces. However, there will likely be some areas where the Automated forecast is less accurate than the existing process, or even less accurate than the Naïve forecast. Once these areas are identified, they can be investigated. (It is important to note that for some sales patterns, the naïve “no change” model is the most appropriate forecasting model, and cannot be meaningfully improved upon.)   To make it easy to identify non-value adding areas, you can build a simple application using SAS® Visual Analytics software. Such an application lets you point and click your way through the organization’s forecasting hierarchy, and at each point view performance of the Naïve, Manual, Statistical, and Automated forecasts (or whatever different types of forecasting methods you are using). This makes it very quick and easy to identify where forecasting methods are not adding value, so these areas can be investigated. SAS Visual Analytics provides a much better environment for doing this kind of analysis than trying to do it in Excel.    The value of an automated (or largely automated) forecasting process is threefold:   It can significantly reduce the amount of time spent on manual forecasting. This frees those resources currently engaged in forecasting to spend more time on sales and customer service activities that can increase revenue and customer satisfaction. Analysts can focus their efforts on the most important, or most problematic forecasts, and let the rest run on auto-pilot. Automated forecasting will generally create more accurate forecasts. In many areas accuracy will be significantly better than the Naïve forecast, although sometimes accuracy will just be comparable to the Naïve. In areas where Manual forecasting does well, because the forecasters have information not available (or not suitable) for computer modeling, then manual adjustments can still be made to the statistical forecast. But automatic forecasting greatly reduces the need for manual efforts – only requiring them in areas where we have determined they are adding value. Automated forecasting should create unbiased forecasts. Manual forecasting processes often have significant bias to over-forecast (forecasts higher than the actual sales turn out to be, due to wishful thinking, or to drive higher inventory and not leave unfilled orders). This adds costs in excess / obsolete capacity and supply. Some organizations have a bias to under-forecast, which might lower inventory costs, at the risk of poor customer service. Automated statistical forecasts should be virtually unbiased – neither chronically too low or too high compared to the actual sales.   In large-scale forecasting situations, automated forecasting is not just an option, it may be a necessity. It is not uncommon for a retailer, having thousands of items sold at hundreds of stores, to have over a million store/item forecasts they want to create. But no company, in any industry, can afford the army of analysts needed to manually build a million+ forecasting models.   When you combine FVA analysis with the large-scale automated forecasting capabilities in SAS Forecast Server (and larger-scale capabilities when you also utilize SAS® Grid Manager), even the largest enterprise can efficiently generate quality forecasts, and focus analyst efforts where they are most needed. ... View more

For the fourteenth year, the International Institute of Forecasters, in collaboration with SAS®, is proud to announce financial support for research on how to improve forecasting methods and business forecasting practice. The award for the 2016-2017 year will be two $5,000 grants, in Business Applications and Methodology.   Criteria for the award of the grant will include likely impact on forecasting methods and business applications. Consideration will be given to new researchers in the field and whether supplementary funding is possible.   Applications must include: Description of the project (max. 4 pages) Letter of support from the home institution where the researcher is based. Brief (max. 4 page) c.v. Budget and work-plan for the project. The deadline for applications is September 30, 2016.  For a complete overview of the requirements for the award, click here.   All applications or inquiries should be sent to IIF Business Director (pamstroud@forecasters.org).   The IIF-SAS grant was created in 2002 by the IIF, with financial support from the SAS Institute, in order to promote research on forecasting principles and practice. The fund provided amounts of US $10,000 per year, which is divided to support research in the two basic aspects of forecasting: development of theoretical results and new methods, and practical applications with real-world comparisons. ... View more

For the thirteenth year, the IIF, in collaboration with SAS®, is proud to announce financial support for research on how to improve forecasting methods and business forecasting practice. The award for the 2015-2016 year will be (2) $5,000 grants. The deadline for applications is September 30, 2015.  For the full description and requirements for the award, click here.   Applications must include: Description of the project (max. 4 pages) Letter of support from the home institution where the researcher is based. Brief (max. 4 page) c.v. Budget and work-plan for the project. All applications or inquiries should be sent to IIF Business Director (forecasters@forecasters.org).   The IIF-SAS grant was created in 2002 by the IIF, with financial support from the SAS® Institute, in order to promote research on forecasting principles and practice. The fund provided amounts of US $10,000 per year, which is divided to support research in the two basic aspects of forecasting: development of theoretical results and new methods, and practical applications with real-world comparisons.   Full details are available at: Grants and Research Awards | International Institute of Forecasters ... View more