24
P-1 © 2005 NeuralWare. All rights reserved . Using Neural Networks in Decision Support Using Neural Networks in Decision Support Systems Systems Introduction Introduction Core Core Technology Technology Building and Building and Deploying Deploying Neural Neural Networks Networks Medical Medical Procedure Procedure Certificatio Certificatio n Crime Crime Forecasting Forecasting Grain Grain Quality Quality Assessment Assessment Jack Copper NeuralWare [email protected] Hiroshi Maruyama SET Software Co. Ltd. [email protected] April 2005

P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

Embed Size (px)

Citation preview

Page 1: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-1© 2005 NeuralWare. All rights reserved.

Using Neural Networks in Decision Support SystemsUsing Neural Networks in Decision Support SystemsUsing Neural Networks in Decision Support SystemsUsing Neural Networks in Decision Support Systems

IntroductionIntroduction

Core Core TechnologyTechnology

Building and Building and Deploying Deploying Neural Neural NetworksNetworks

Medical Medical Procedure Procedure CertificationCertification

Crime Crime ForecastingForecasting

Grain Quality Grain Quality Assessment Assessment

Jack Copper

NeuralWare

[email protected]

Hiroshi Maruyama

SET Software Co. Ltd.

[email protected]

April 2005

Page 2: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-2© 2005 NeuralWare. All rights reserved.

NeuralWareNeuralWareNeuralWareNeuralWare

Since 1987, NeuralWare has created and marketed neural network based Artificial Intelligence (AI) software for –

Data Mining (clustering) Classification Forecasting

NeuralWare collaborates with Customers and Partners to Embed Intelligent Neural Network Engines into Next-Generation Products and Systems

Page 3: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-3© 2005 NeuralWare. All rights reserved.

IntroductionIntroductionIntroductionIntroduction

Characteristics of Neural Network Decision Support Systems

Integrate Data and Analytics

Adapt to Changing Conditions

Page 4: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-4© 2005 NeuralWare. All rights reserved.

IntroductionIntroductionIntroductionIntroduction

Benefits of Neural Network Decision Support Systems

Consistent Decisions

Rapid Decisions

Reproducible Decisions

Page 5: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-5© 2005 NeuralWare. All rights reserved.

IntroductionIntroductionIntroductionIntroduction

Examples of Neural Network Decision Support Systems

Medical Procedure Certification

Crime Forecasting

Grain Quality Assessments

Page 6: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-6© 2005 NeuralWare. All rights reserved.

Core Technology - Neural NetworksCore Technology - Neural NetworksCore Technology - Neural NetworksCore Technology - Neural Networks

Historic Data

Target

Model

Target

Model

Input LayerInput Layer

Hidden LayerHidden Layer

Output LayerOutput Layer Decisions Based on Model OutputDecisions Based on Model Output

New DataNew Data

Artificial Neural Networks are connected hierarchies Artificial Neural Networks are connected hierarchies of Artificial Neurons (also called Processing Elements)of Artificial Neurons (also called Processing Elements)

Page 7: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-7© 2005 NeuralWare. All rights reserved.

Building Neural NetworksBuilding Neural NetworksBuilding Neural NetworksBuilding Neural Networks

Page 8: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-8© 2005 NeuralWare. All rights reserved.

Evaluating Neural Network PerformanceEvaluating Neural Network PerformanceEvaluating Neural Network PerformanceEvaluating Neural Network Performance

Page 9: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-9© 2005 NeuralWare. All rights reserved.

Evaluating Neural Network PerformanceEvaluating Neural Network PerformanceEvaluating Neural Network PerformanceEvaluating Neural Network Performance

Page 10: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-10© 2005 NeuralWare. All rights reserved.

Deploying Neural NetworksDeploying Neural NetworksDeploying Neural NetworksDeploying Neural Networks

NeuralWare Technology (Run-Time Engine/Models/FlashCode) embedded in Server

Browser-based wired or wireless remote PC clients do not employ NeuralWare technology

Server Contains Development and Run-Time Engine

Application Server Architecture

Page 11: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-11© 2005 NeuralWare. All rights reserved.

Deploying Neural NetworksDeploying Neural NetworksDeploying Neural NetworksDeploying Neural Networks

Wired or wireless remote PC clients employ embedded NeuralWare technology (Run-Time Engine/Models/FlashCode)

Server Contains Development Engine

Distributed Intelligence Architecture

Page 12: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-12© 2005 NeuralWare. All rights reserved.

Case Study – Medical Procedure CertificationCase Study – Medical Procedure CertificationCase Study – Medical Procedure CertificationCase Study – Medical Procedure Certification

Objectives

Reduce Workload on Doctors and Registered Nurses

Improve Responsiveness to Customers (faster decisions)

Challenges

No “Gold Standard” for decisions – even Doctors sometimes disagree

Inconsistent data formats and labeling

Process

Used NeuralSight to build and evaluate ~ 30,000 Models in 3 weeks

Developed prototype software to permit altering Model decision threshold

Page 13: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-13© 2005 NeuralWare. All rights reserved.

Case Study – Medical Procedure CertificationCase Study – Medical Procedure CertificationCase Study – Medical Procedure CertificationCase Study – Medical Procedure Certification

Performance of best models (ranked by Average Classification Rate) for the Global model and CT and MRI Modality models

Page 14: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-14© 2005 NeuralWare. All rights reserved.

Case Study – Medical Procedure CertificationCase Study – Medical Procedure CertificationCase Study – Medical Procedure CertificationCase Study – Medical Procedure Certification

Page 15: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-15© 2005 NeuralWare. All rights reserved.

Case Study – Medical Procedure CertificationCase Study – Medical Procedure CertificationCase Study – Medical Procedure CertificationCase Study – Medical Procedure Certification

Acquire/Validate Case Input

Retrieve Metrics

Select/Execute Model

Apply Thresholds

Approve Procedure?

Process ManuallyUpdate Metrics

Selected for Audit?

NO

YES

YES

DONE

Metric Database

Page 16: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-16© 2005 NeuralWare. All rights reserved.

Case Study – Crime ForecastingCase Study – Crime ForecastingCase Study – Crime ForecastingCase Study – Crime Forecasting

Objectives

Identify Patterns in Criminal Activity that indicate Potential Future Trouble Spots

Redirect Police Resources to Focus on Areas where Serious Crime is expected to Increase

Challenges

Defining Crime Categories and Severity Levels

Inconsistent data formats and labeling; missing or non-existent data

Process

Used NeuralSight to Build and Evaluate ~ 10,000 Models in 1 week

On-going evaluation by researchers at Carnegie Mellon University

Page 17: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-17© 2005 NeuralWare. All rights reserved.

Case Study – Crime ForecastingCase Study – Crime ForecastingCase Study – Crime ForecastingCase Study – Crime Forecasting

Page 18: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-18© 2005 NeuralWare. All rights reserved.

Case Study – Crime ForecastingCase Study – Crime ForecastingCase Study – Crime ForecastingCase Study – Crime Forecasting

How to Forecast Change in Crime

Police know current crime levels Have allocated resources to respond to existing crimes

Most valuable information for tactical level planning: Where is crime likely to have large increases next month? Forecast crime by area and calculate:

Forecasted Change (t+1) = Forecast (t+1) – Actual (t)

The Benefit – Better Allocation of Scarce Resources

Page 19: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-19© 2005 NeuralWare. All rights reserved.

Case Study – Crime ForecastingCase Study – Crime ForecastingCase Study – Crime ForecastingCase Study – Crime Forecasting

Forecasted Change for July

Page 20: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-20© 2005 NeuralWare. All rights reserved.

Case Study – Grain Quality AssessmentCase Study – Grain Quality AssessmentCase Study – Grain Quality AssessmentCase Study – Grain Quality Assessment

Objectives

Provide a Platform for rapidly and consistently assessing the quality of grain

Maintain detailed records of tests and build foundation for data mining

Challenges

No “Gold Standard” for decisions – even experienced human inspectors are inconsistent

Requires tedious work to identify wide variety of training data samples

Process

Used Predict and NeuralSight to Build and Evaluate many thousands of Models

Now developing image database to support agriculture research

Page 21: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-21© 2005 NeuralWare. All rights reserved.

Case Study – Grain Quality AssessmentCase Study – Grain Quality AssessmentCase Study – Grain Quality AssessmentCase Study – Grain Quality Assessment

An Instrument – and examples of seed images

Page 22: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-22© 2005 NeuralWare. All rights reserved.

Case Study – Grain Quality AssessmentCase Study – Grain Quality AssessmentCase Study – Grain Quality AssessmentCase Study – Grain Quality Assessment

Many (more than 300) initial features per seed Predict Variable Selection found a much smaller set of

features to use in building models

The characteristics of grain that are important are difficult even for human inspectors to identify Multiple neural networks are used to make the hard decisions

The value of wheat and other commodities depends on its quality – millions of dollars are at risk if quality decisions are incorrect!

Page 23: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-23© 2005 NeuralWare. All rights reserved.

What Have you Learned?What Have you Learned?What Have you Learned?What Have you Learned?

Neural Networks make Powerful Decision Support Systems Human Judgment Determines the Cost/Benefit Tradeoff for Accuracy

Know your Problem !

Neural Network Decisions are based on Learning Patterns Relationships in Historical Data are the basis for Current Action

Know your Data !

Page 24: P-1 © 2005 NeuralWare. All rights reserved. Using Neural Networks in Decision Support Systems Introduction Core Technology Building and Deploying Neural

P-24© 2005 NeuralWare. All rights reserved.

Thank You !Thank You !Thank You !Thank You !

Jack Copper

NeuralWare

[email protected]

Hiroshi Maruyama

SET Software Co. Ltd.

[email protected]