Visual Numerics’ Neural Networks forecasting classes can help users discover relationships
and valuable information in vast amounts of data. Neural Networks provide a high-flexibility
forecasting approach that delivers accurate forecasts under a variety of conditions including
complex data systems, short time series, mixed categorical and continuous data, noisy data
and large numbers of variables.

Visual Numerics’ Neural Network technology features advanced data pre-processing to
simplify data mining preparation. These automatic pre-processing algorithms are tailored to
work together with the Neural Network training engine to ensure optimal accuracy and
performance, saving the user significant time over manual pre-processing. Unlike other
neural network algorithms, the Neural Network training algorithm implementation in the
IMSL Family allows the user control over parameters such as the number of training epochs,
specific linkages within the network and activation functions. In addition, there is no limit to
the number of layers and perceptions that can be used in the network. This high degree of
flexibility and control is especially beneficial to analysts in the fields of finance, business
analytics, bioinformatics and life sciences, where data sets can be extremely large and
complex.

Teksouth: Advanced Predictive Analytics Solution For More Accurate
Budget Forecasting
With demanding customers, Teksouth wanted to go beyond traditional forecasting tools and
into true business intelligence in order to help customers more accurately project and budget
spending changes.

In partnering with Teksouth to develop a Neural Network forecasting application, Visual
Numerics tapped into its mathematical and statistical libraries. The Neural Network
forecasting application extracts information from historical cost data, applying the
information to forecast future costs. The analytic engine will have the flexibility to prioritize
historical data and exclude anomalous data to improve its predictive capabilities. The
solution applies knowledge gained form historical data to new problems, fine-tuning its
forecasting accuracy even more over time.

"The flexibility and learning potential of our neural net forecasting application has significant potential to help government agencies and commercial businesses improve their predictive
analysis capabilities, moving well beyond traditional tools into high end business intelligence."

-Larry Northington vice president of strategic planning and corporate development for Teksouth, and a retired u.s. air force major general.

Major European Grocery Chain
In the very competitive grocery industry, companies win or lose based on their ability to predict customer shopping habits. A major grocery chain in Europe that handles hundreds of thousands of products was relying on historical average order statistics to forecast product demand and inventory. For some products, these statistics did not provide forecasts of short term inventory requirements. They wanted to do better. The grocery chain turned to Visual Numerics Professional Services to help them use more advanced statistical and data mining techniques for predicting inventory requirements. They wanted to quickly and accurately forecast the demand for each product for each of their stores. Visual Numerics is delivering a total data mining solution through its extensive libraries, in-house data mining expertise, and development resources, packaging this technology as a customized, configurable and easy to use solution. Neural Net preprocessing, forecasting and training engines will be provided for integration into the customer’s current enterprise .NET applications used by store managers.

The Auto_ARIMA Advanced Forecasting Routine for Time Series Analysis
The IMSL Library function, Auto_ARIMA, is an advanced forecasting routine for time series
analysis with an ARIMA auto modeler. It is named using the term "auto," as in automatic, since
calling the function requires very little, if any, pre-processing of the time series data by the
user, unlike traditional methods.

ARIMA is a powerful and flexible methodology, but it is an extremely complex technique and
requires a great deal of experience to be used effectively. The accuracy of results heavily
depends on the analyst's level of expertise. With Auto_ARIMA, the complexity of time series
analysis with ARIMA is greatly simplified due to its accurate and automatic pre-processing of
data, allowing for faster time to analysis.

The Auto_ARIMA function is applicable for applications such as sales forecasting, commodity
pricing (i.e. oil & gas), stock market predictions, semiconductor yield analysis, and more.
Utilizing the expertise of Visual Numerics’ Professional Services Team, an organization can
integrate Auto_ARIMA to create the optimal forecasting solution for their application.

Major European Telecommunications Company: Advanced Forecasting
For Cellular Network Optimization
For a major European mobile telecommunications company, planning and managing their
cellular network is key to their success. Too little coverage and customers experience service
problems; too much coverage and resources are wasted. Good planning means analyzing the
massive amounts of usage data generated by the thousands of cellular towers in the service
provider’s network. To track usage on their German mobile network, the service provider already utilized sophisticated data analysis techniques, but needed another level of refinement in order to better understand traffic patterns.

For a company owning thousands of cellular towers, all sending data to their Network Operations Center, traditional statistical analysis techniques were inadequate, labor intensive, expensive and cumbersome. The service provider turned to Visual Numerics Professional Services group. To identify the outliers and locate any seasonal trends in the data, Visual Numerics leveraged the ARIMA algorithm from the IMSL Numerical Library. Visual Numerics’ software engineers then automated the algorithm to tackle the customer’s faster processing requirement. The state-of-the-art, expert-system Auto_ARIMA algorithm allowed the service provider to produce more accurate forecasts faster than before, giving them a significant advantage in the market.

Decades of Experience for Maximum Value
Visual Numerics’ IMSL Numerical Libraries and PV-WAVE visual data analysis tools have
been the cornerstone of high-performance and deep computing as well as predictive analytics
applications in science, technical and business environments for well over three decades.
These tools are regarded as the most sophisticated, flexible, scalable and highly accessible
technology available for advanced analytics in the most important mainstream programming
environments in use today.

Key Benefits

• Optimized and supported algorithms
• Unparalled scalabilty and interoperability
• Broad range of solutions for classic and advanced forecasting problems
• Over three decades of expertise in the area of forecasting
• Expert Professional Services for custom algorithm and application development
• Unmatched product support
• Seamless and simplified integration
• Time-tested technology for more accurate result