Mining data with Poly Analyst learn to profit

Mining data with Poly. Analyst learn to profit from data © 1999 Megaputer intelligence, Inc.

Outline O Why data mining? O Poly. Analyst solution overview O Learning algorithms O Data handling and result reporting O Customer cases O Future developments

Problem Here is aabusiness chain: Consider fragmentinformation of the chain: Data Knowledge Decision Action O Data - is what we can capture and store O Knowledge - is what provides for informed decisions O Problem: How to get Knowledge from data? Data Knowledge

Solution O Computers can be taught to learn from data O Data mining is a process of discovering new and valuable knowledge hidden in data O Poly. Analyst is the most comprehensive and user-friendly data mining solution for PC Poly. Analyst Data Knowledge

Poly. Analyst overview

What is Poly. Analyst? Poly. Analyst O multi-strategy data mining suite O utilizing the latest achievements in knowledge discovery in databases O powerful data manipulation and visualization tools

Data Access O Poly. Analyst works with ODBC-compliant databases: Oracle, DB 2, Informix, Sybase, MS SQL Server, etc. O Data and exploration results can be exchanged with MS Excel O CSV format files

Tasks solved O Modeling O Predicting O Clustering O Classifying O Explaining

Poly. Analyst provides O Reliable access to data held in a database or data warehouse (works with numerical, categorical, and yes/no variables) O Comprehensive functions for data manipulation and visualization O Multiple machine learning algorithms O Convenient result reporting, printing, and incorporating in a decision support system

Poly. Analyst workplace Multiple machine learning algorithms can be accessed through pull-down and pop-up menus, or control buttons The project data, charts, discovered rules, and system reports are represented by icons held in separate containers

Fields of application

Poly. Analyst machine learning algorithms

Learning algorithms O Find Laws (SKAT algorithm) O Cluster (Localization of anomalies) O Find Dependencies (n-dimensional distributions) O Classify (Fuzzy logic modeling) O Poly. Net Predictor (GMDH-Neural Net hybrid) O Linear Regression (Stepwise and rule-enriched) And available in the new Poly. Analyst COM: O Memory Based Reasoning (k-Nearest Neighbor)

Poly. Analyst Customer base: 100+ installations Poly. Analyst key customers: O Boeing (USA) O 3 M (USA) O Alka Insurance (Denmark) O Carlson Marketing (USA) O France Telecom (France) O Allstate Insurance (USA) O University of Ulm (Germany) O Pioneer Securities (Russia)

Sample customer cases

Poly. Analyst supports medical projects at 3 M Timothy Nagle Consulting Scientist 3 M Corporation St. Paul, MN, USA “. I found the user interface both intuitive and easy to use. Megaputer support is outstanding. The inevitable problems one expects with a complex system are dealt with immediately. ”

Poly. Analyst helps improving flight control system at Boeing James Farkas Senior Navigation Engineer The Boeing Company Kent, WA, USA “Poly. Analyst provides quick and easy access for inexperienced users to powerful modeling tools. Interfaces to spreadsheet tools provide flexibility needed to work solutions as a team. ”

Poly. Analyst Future Developments

Future developments O New machine learning algorithms: – Memory Based Reasoning – Weighted variable Clustering and Classification O Poly. Analyst COM built on the basis of Component Object Model - an integrated kit for simple development of decision support applications utilizing advanced Poly. Analyst algorithms (see PCAI Magazine, March 99, p. 16 -19) O Enhanced graphics (Snake and Boxplot charts) and data import and manipulation

Poly. Analyst evaluation We invite you to download an evaluation copy of Poly. Analyst from www. megaputer. com and enjoy using it hands-on following provided step-by-step lessons, or using your own data. Megaputer team looks forward to helping you with your project! © 1999 Megaputer intelligence, Inc.