Adaptive Business Intelligence by Zbigniew Michalewicz

By Zbigniew Michalewicz

Within the glossy info period, managers needs to realize the aggressive possibilities represented via decision-support instruments. Adaptive enterprise intelligence structures mix prediction and optimization suggestions to help selection makers in advanced, quickly altering environments. those platforms tackle the elemental questions: what's prone to ensue sooner or later? and what's the simplest determination at once? Adaptive company Intelligence contains components of knowledge mining, predictive modeling, forecasting, optimization, and adaptability.The authors have massive educational learn backgrounds in synthetic intelligence and similar fields, mixed with years of functional consulting adventure in companies and industries world wide. during this ebook they clarify the technological know-how and alertness of diverse prediction and optimization suggestions, in addition to how those strategies can be utilized to enhance adaptive platforms. The strategies coated comprise linear regression, time-series forecasting, choice bushes and tables, synthetic neural networks, genetic programming, fuzzy structures, genetic algorithms, simulated annealing, tabu seek, ant platforms, and agent-based modeling.This publication is acceptable for company and IT managers who make judgements in complicated business and repair environments, nonspecialists who are looking to comprehend the technological know-how at the back of greater predictions and judgements, and scholars and researchers who want a speedy creation to this box.

Show description

Read Online or Download Adaptive Business Intelligence PDF

Best management information systems books

Enabling Optical Internet with Advanced Network Technologies

Advancements in optical transmission platforms and digital routers permit large bandwidth and ever-increasing ability. but the new gigantic deployment of ADSL and broadband entry, including the outburst of latest multimedia community providers, have led to an important raise in end-user site visitors and bandwidth calls for.

Building a future with BRICs: the next decade for offshoring

In 2003, Goldman Sachs released a startling document at the BRICs (Brazil, Russia, India, and China) quarter: those 4 nations will be better than the G6 economics inside of forty years, muscling their approach to monetary dominance and powering prior built international locations reminiscent of the uk, Germany, and Japan.

Introduction to Distributed Algorithms

The cheaper price and elevated flexibility and availability of computing device networks implies that they're swiftly exchanging centralized 'mainframes'. The implementation of networked platforms and the keep watch over of purposes working in them require, besides the fact that, the layout of recent ideas and strategies: jointly often called disbursed algorithms.

Pragmatic Enterprise Architecture Strategies to Transform Information Systems in the Era of Big Data

Pragmatic firm Architecture is a realistic hands-on guide for firm architects. This ebook prepares you to higher interact IT, administration, and company clients via equipping you with the instruments and information you want to tackle the commonest company structure demanding situations. you'll come away with a practical knowing of and method of company structure and actionable rules to remodel your online business.

Extra info for Adaptive Business Intelligence

Sample text

If the adaptability module updates the prediction module every week by using a rolling time window, then the prediction module can adapt to changes in the sale prices. Alternatively, imagine that the prediction module has certain rules that can be expressed as follows: if [Make = Honda] and [Model = Accord] and [Color = white] and [40,000 < Mileage < 50,000] and [Year = 2000] and [Damage Level = $0], then Sale Price = $11,384. 5 The Structure of an Adaptive Business Intelligence System 45 Each of these rules has a weight,15 and the weights of rules can be modified (say, on a weekly basis) to tune the predictions in a changing environment.

3 Optimization 43 Let us also assume that our (very basic) prediction module “predicts” the sale price for each make/model at each auction site by looking at the average sale prices for the previous week (hence, it does not take into account the actual mileage, color, or other variables). It is interesting to note, however, that creating even a basic prediction module such as this presents many difficulties. For instance, a separate prediction model is needed for each make, model, and location.

New transactions) and the number of variables in each case (e. , the detail of each transaction). In general, the more data the better, as data mining can produce better results when performed on large data sets, and the resulting prediction models are more accurate. In the car distribution example, there are several important elements of prediction. For example, we would like to predict the sale prices for different cars at different auction sites on different days. Because these predictions are based on past cases, we should know all the variables (e.

Download PDF sample

Rated 4.91 of 5 – based on 24 votes