Intellia Logo Home
Intellia (Auckland, New Zealand)
 
mm_spacer.gif
mm_spacer.gif
     
mm_spacer.gif
mm_spacer.gif
mm_spacer.gif
 

Artificial Neural Networks
Neural Network Software Design
 
Neural Network Software: Price Forecasting Neural Network Software: Credit Approval
(Artificial Intelligence) (Data Mining) (Decision Support) (Artificial Intelligence) (Data Mining) (Decision Support)

Neural Network for Price Forecasting (Intellia Ltd)

Neural Network for Credit Approval (Intellia Ltd)

 



Image above: Neural Network Interface for Price Forecasting

The image above shows part of the calculations interface that displays the neural networks calculations of the transform data, sigmoid calculations, thresholds, backpropagations, iterations in progress, and the validated results, based on the neural networks training. Neural network training is usually conducted using an expert that is familiar with the mathematics and operation of neural networks, and who understands the business basics the neural network is being applied to.

The Price Forecasting Neural Network uses 5 input layers whose
price data inputs consists of a 5-days moving average, a 10-days moving average, a 20-days moving average, a 60-days moving average and a 120-days moving average. In addition, 3 hidden layers are applied to the input data, to generate the desired output which is the prediction of next weeks price for each item.


For one particular exercise conducted by a business consultancy for a corporate client, this particular neural network undertook close to 10,000 iterations on price data, before the neural network arrived at its optimum training level. Iteration is the act of repeating a process with the aim of approaching a desired goal, target or result. Each repetition of the process is an iteration, and the results of one iteration are used as the starting point for the next iteration. Therefore to determine the optimum price forecast the neural network parsed, calculated and processed the data almost 10,000 times, before achieving the optimum price forecast.

The Credit Approval Neural Network, contains 10 data input layers,
6 hidden layers, and 1 output layer. The output layer specifies all
the credit applications that should be approved and all the credit applications that should be declined.


About Artificial Neural Networks


Neural networks are a decision support technology used to support operational control, managerial control, and strategic planning control, where the company deals with unstructured type of problems. Unstructured problems are problems for which no standard solution methods exist; where processes are fuzzy, and problems are too complex to apply structured decision methods to.

Neural Network Software helps identify patterns and profiles in data
sets to help provide information for decision-making. Neural Networks need to be customized to correspond to the data being analyzed and neural networks need to be trained for the type of decisions that need to be made.

Neural networks can often succeed where TPS, MIS, DSS, KMS, IDSS, ES, EIS, SCM and human decision makers would fail. Artificial Neural Networks (ANN) can be used for many business applications across business departements and industry sectors.

Available Neural Network Software:

  • Neural Network for Price Forecasting
  • Neural Network for Credit Approval
  • Neural Network for Sales Forecasting
  • Neural Network for Object Classification
  • Neural Network for Real Estate Value Prediction
  • Neural Network for Stock Market Prediction

 

   
Consult with us about neural network business applications that provide the industrial application to solve unstructured problems to provide decision support for your organisation.

businessanalysis@intellia.co.nz
 
   
   
   
 
 

COPYRIGHT NOTICE:

Intellia (New Zealand) Ltd software designs and software products are the intellectual property and copyright of Intellia Ltd, Auckland, New Zealand. Software titles are designed, developed and programmed by Stefan Reucker (director). Many of these applications, though they conform in some cases, at least in part, to established design patterns, are software titles that are used as part of a consultation process between Intellia and it's consulting business clients; they are working software product models that are designed to be customizable to meet the specific business clients particular business needs and they are thus used either as is or as the basis for future customization and additional software development to enhance the software prototype. These software titles are therefore used either as off-the-shelf products to be used right away, or as a foundation for further business analysis and custom software engineering. The consultant-to-business relationship ensures that the business client receives what the specific business needs and that the business client has software that is truly software designed for that particular business, taking into consideration aspects of public i.e. governmental organizations accountability, privacy and ownership. The software is compliant with New Zealand legislation and international norms.

     
 

 

| About Us | Consulting Process | Contact Us | Privacy Policy | Sitemap | Terms & Conditions |

Logo of Intellia.png

Copyright © Intellia Ltd, 2008 - 2016.

Auckland CBD, New Zealand.