Announcing StatTools and
NeuralTools for Excel 2007
StatTools and NeuralTools are now fully compatible with Excel 2007. Existing users with current maintenance plans can get a FREE update for their software.

Take advantage of the new features in Excel 2007, including:

Expanded Worksheet – Excel 2007 offers 1 million rows and 16,000 columns in a single worksheet, allowing for vastly larger model sizes and data.

Improved Ribbon interface – StatTools and NeuralTools commands and features are now easier to find on the Excel Ribbon, which visually brings together the essentials.

Improved Charting - Graphs created by StatTools are fully customizable using all of Excel 2007's new graphing features. In addition, Excel 2007 charts are more easily shared with other Office applications such as Word and PowerPoint.

» Get your StatTools update
» Get your NeuralTools update

NeuralTools Used to Speed
Prediction Time for Cardiovascular
Risk Assessment
Source: Scientific Computing, July 2007
The selection and dosing of drugs is an important component of many care plans designed by clinicians in a wide variety of healthcare settings. Researchers used Palisade’s NeuralTools to evaluate factors that influence both selection and dosing of drugs based upon pathophysiological and pharmacological principles.

» Read the full story
» Learn more about NeuralTools

Free Live Webcast:
How to Measure Anything: Finding
the Value of Intangibles in Business
Learn how to make intangibles measurable thus improving your models in this FREE live webcast on Palisade.com. Presented by Douglas W. Hubbard, President, Hubbard Decision Research

Doug Hubbard will talk about the approach described in his book, How to Measure Anything: Finding the Value of Intangibles. This is based on Hubbard’s method Applied Information Economics (AIE) and includes how to formulate intangibles as measurable and techniques for improving Monte Carlo models including:

1. How “calibration training” improves on the subjective estimates from subject matter experts.
2. How to compute the value of additional information and how this radically changes empirical measurement methods.
3. How to use the output of a Monte Carlo simulation as input to portfolio optimization methods.

» Register now for FREE Live Webcast:
"How to Measure Anything:
Finding the Value of Intangibles in Business"
23 August 2007, 11:00am - 12:00pm ET / 4:00 - 5:00 pm GMT

@RISK Combats Avian Flu
The Royal Veterinary College (RVC), the UK’s largest vet school and one of the most highly regarded veterinary research centers in Europe, uses @RISK to augment its epidemiology skills when understanding the risks associated with specific diseases. In the case of avian flu the RVC is able to assist in containing something that is widely perceived as highly threatening.

» Read the full case study
» Learn more about @RISK

@RISK Stress Analysis
@RISK Stress Analysis lets you control the range that is sampled from a distribution function, enabling you to see how different scenarios affect your bottom line without changing your model. By controlling the values sampled from a distribution, you stress the distribution. Simply specify the percentile values between which samples are drawn during simulation. Stress Analysis also allows you to stress your model by substituting an entirely new distribution for an existing distribution in your model. Instantly test various scenarios without changing your model! Stress Analysis is included with the Professional and Industrial editions of @RISK 4.5.

As an example, an insurance company might want to analyze its loss reserves. It would like to see what the effect would be on claims payout if it took out a reinsurance policy to cover an extraordinary number of claims. By stressing the distribution representing the number of claims to represent just the 90th percentile and below, you could see if the reserves on hand are enough to cover 90% of possible claims filed, while covering the highest 10% with an outside policy.

After completing a simulation, Stress Analysis provides you with a collection of reports and graphs you can use to analyze the effects. These include:

• Summary reports
• Box-Whisker plots
• Comparison graphs
• Histograms
• And more

» More about @RISK

Atomic Movement of Solids

» DC.Ars: Team Heavy Metal Lasagna
and the eOn Project
Technique: Monte Carlo simulation  » see @RISK
Source: Ars Technica, 20 July 2007

The goal of the EON project is to calculate the growth/movement of solids at the atomic scale via Monte Carlo simulations. This simulation method has been employed to study island formation and growth, and the EON project is devoted to the growth of ice.

Biological Research

» Learning to evolve: With a little help
from my ancestors
Technique: Neural networks / Genetic algorithms   » see NeuralTools
Source: PhysOrg.com, 30 July 2007

According to a new theory, the innate ability of today's birds to fly depends indirectly on the learning that their ancestors did, which leaves a genetically specified latent memory for flying. The theory has been tested on simple models of brains called artificial neural networks, which can be made to evolve using genetic algorithms.

Cancer Treatment

» Monte Carlo Dose Calculation Extends CyberKnife(R) System's Capabilities in Precision Radiation Delivery
Technique: Monte Carlo simulation  » see @RISK
Source: CNNMoney.com, 23 July 2007

Accuray Incorporated, a global leader in the field of radiosurgery, announced an extension of the CyberKnife System's suite of products targeted toward the noninvasive treatment of lung tumors with Accuray's Monte Carlo dose calculation algorithm, which is pending FDA 510(k) clearance.

Credit Analysis

» Fitch Expects to Rate Metrofinanciera Trust #650 - Construction Bridge Loan Securitization
Technique: Monte Carlo simulation  » see @RISK
Source: Business Wire, 18 July 2007

Fitch focuses credit analysis on several key quantitative and qualitative variables that could affect the performance of the underlying assets. Fitch estimates default risk on the portfolio using a a multi-period Monte Carlo simulation model, which simulates the default behavior of an individual asset for each year of the transaction's life. The Monte Carlo simulation allows for the modeling of the distribution of portfolio defaults and losses, taking into account the correlation between assets in a portfolio.

Credit Card Fraud

» The new fraud squad
Technique: Neural networks   » see NeuralTools
Source: VNUnet.com, 26 July 2007

Costing £212.6m each year in the UK and growing at an annual rate of 16 per cent, cardholder-not-present (CNP) fraud is now the biggest category of plastic card deception. The internet retail industry in the UK is growing at an exceptional rate and credit card fraudsters are attacking the softer targets of the online sector. One category of fraud prevention systems is neural networks that can be built into payment processing systems in order to identify potentially deceitful transactions.

Currency Forecasting

» Propipper.com Offers Forecasts
Utilizing Neural Networks
Technique: Neural networks   » see NeuralTools
Source: PR-GB.com, 13 July 2007

As part of the services, Pro Pipper will offer The New Wave Forex Market Watch service, including Inter-market Analysis utilizing neural networks to analyze correlation patterns between the target currency and its related commodity, bond and stock index markets.

Diabetes Treatment

» Top US, EU diabetes groups offer guidelines for Asians
Technique: Decision trees  » see DecisionTools Suite
Source: Inquirer.net, 28 July 2007

The rapidly rising prevalence of diabetes among Asian nations has led experts from the American Diabetes Association to be alarmed and visit the Asia-Pacific region, for the first time, to offer their recommendations -- a treatment algorithm -- for the management of type 2 diabetes. A treatment algorithm usually refers to decision tree approaches to healthcare treatment (i.e., if symptoms A, B and C are evident, then use treatment X).

Disease Diagnosis

» About.com Symptom Checker
Technique: Decision trees  » see DecisionTools Suite
Source: Business Wire, 31 July 2007

About.com has launched a new symptom checker, an interactive tool developed in conjunction with Harvard Medical School. The tool covers 100+ common symptoms and guides users through a decision tree to determine possible conditions and appropriate actions. On average, 34 million monthly visitors use About.com to help guide their decisions about Health, Finance and many other topics.

Financial Planning

» Rating the Online Retirement Planners
Technique: Monte Carlo simulation  » see @RISK
Source: Kiplinger’s Personal Finance, August 2007

Kiplinger’s rates the top online retirement planning services, finding the more advanced tools incorporate Monte Carlo simulation into their investment calculations.

» John Hancock Funds Adds New Retirement
Portfolio Calculator using Monte Carlo simulation
Technique: Monte Carlo simulation  » see @RISK
Source: PR Newswire, 1 August 2007

John Hancock Funds recently launched a new retirement portfolio calculator on its award-winning web site. The calculator represents the probability that such a portfolio would last throughout retirement, given different retirement lengths and annual withdrawal rates. The chance of portfolio survival was projected using Monte Carlo simulation, which is useful for modeling outcomes with significant uncertainty and variety of inputs.

Food Regulation

» Government Chemist designs system to
classify dietary food and health supplements
for applicable legislation
Technique: Decision trees  » see DecisionTools Suite
Source: Oxfordshire Bioscience Network, 11 July 2007

The creation of a simple system to assign the appropriate legislation for a range of dietary supplements is one of the key achievements described in the Government Chemist Review 2006, the latest account of analytical, advisory and strategic work for the UK Department for Innovation, Universities and Skills (DIUS). A decision tree was created to address the key factors in deciding whether a product falls under the provisions of the Food Safety Act 1990 or the Medicines Act 1968.

» New rules give food vendors leeway;
Health ministry regulations allow for choices
Technique: Decision trees  » see DecisionTools Suite
Source: Kingston Whig-Standard, 11 July 2007

New Ontario provincial regulations allow the local medical officer of health to approve new foods for vendors to sell if they meet safety regulations. For that, the government officials have developed a decision tree, rather than a whole list, to help make good decisions about whether a food could be allowed or not.

Recruiting

» CV Parsing – what is it and why do you need it?
Technique: Neural networks   » see NeuralTools
Source: Onrec.com, 24 July 2007

CV parsing involves the deployment of neural network technology to identify candidate data directly from the resume with a high degree of precision. The artificial intelligence engine is able to read and understand resumes by making probabilistic determinations based on linguistic context and inference.

Robotics

» A Baby Step for Computer Language Acquisition
Technique: Neural networks   » see NeuralTools
Source: ScienceNOW, 23 July 2007

Neuroscientists studying language and cognition at Stanford University are using neural networks to teacher their computers to recognize and distinguish between vowel sounds, just the way an infant learns to recognize these sounds.

» Dynamic Machine Goes Mountaineering
Technique: Neural networks   » see NeuralTools
Source: Science Daily, 13 July 2007

A team of EU-funded researchers has built a robot that can walk uphill and adjust its gait to traverse ice--or sand. Human changes of gait are triggered by the brain, while the robot’s high level of joint and torso coordination has been achieved by linking an infrared “eye” to a simple neural network.

Securities Valuation

» Qiao Xing Universal Telephone, Inc.
Releases 2006 Annual Report
Technique: Monte Carlo simulation  » see @RISK
Source: PRNewswire, 17 July 2007

Two convertible notes issued in 2006 are hybrid instruments that come with embedded derivatives, including the right to convert the notes by the note holders, a premium put, a put option conditional upon certain events of default and a put option conditional upon a change of control. The valuation of the embedded derivatives was derived by using various valuation methods, which included Monte Carlo Simulation

Supply Chain Optimization

» Genetic Evolution as a Foundation for
Supply Chain Optimization
Technique: Genetic algorithm optimization  » see RISKOptimizer
Source: World Trade Magazine, 30 July 2007

Evolution and other features of the natural world are being used more and more to solve supply chain problems that we haven’t been able to tackle before. In North America, genetic algorithms have been used by many companies since the 1990s to find better solutions to supply chain problems.

» Active vibration control (AVC) of a satellite
boom structure using optimally positioned stacked piezoelectric actuators
Technique: Genetic algorithm optimization
Application: Vibration control of a satellite boom
Source: University of Southampton, 29 July 2007

In this paper, results for active vibration control predicted from experimental measurements on a lightweight structure are compared with purely computational predictions. The structure studied is a 4.5m long satellite boom consisting of 10 identical bays with equilateral triangular cross sections. The study involves a genetic algorithm to find high-quality positions for three actuators on the structure that will achieve the greatest reductions in vibration transmission.

» Estimating Risk of Cancer Associated
With Radiation Exposure From 64-Slice
Computed Tomography Coronary Angiography
Technique: Monte Carlo simulation
Application: Cancer risk analysis
Source: Journal of the American Medical Association, 18 July 2007

Computed tomography coronary angiography (CTCA) has become a common diagnostic test, yet there are few data on its associated cancer risk. A recent report provides a framework based on Monte Carlo simulation for estimating these risks over a lifetime.

» Predicting How Proteins Fold
Technique: Monte Carlo simulation
Application: Drug development
Source: Chemistry World, 6 July 2007

Membrane proteins are important targets for drug development, yet notoriously difficult to study by both experimental and theoretical methods. Researchers in Italy and the UK have now developed a computational approach that can simulate the folding of membrane proteins in atomic detail. The study uses Monte Carlo simulations, which apply millions of random steps to optimize structures for any given environment.

Expert Answers to Technical Questions

Dear Amy,

Can I use Evolver and @RISK to determine the optimal allocations in a portfolio of funds with random growth?

—M.V.

Dear M.V.,

Solving problems of this nature involve two modeling stages. The first stage uses @RISK to simulate the risk in each fund's growth. The second stage uses Evolver to determine the optimal portfolio allocations based on the simulation data generated by @RISK. Typically, the goal of the Evolver optimization would be to maximize the Average Total Return or a specific percentile of the Total Return that represents its VAR (Value at Risk). Additional constraints can be placed on the Total Return statistics like its standard deviation and percentiles.

Example: For demonstration purposes, the underlying @RISK model for generating the individual fund growth data has been simplified in this example. Typically, models that value fund growth subject to risk can be quite complex-often using autoregressive or other ad hoc time-series modeling techniques. These topics are beyond the scope of this article.

» Download this example