Project: {IBA} MINER

Overview

{IBA}Miner is an expert system, developed at the AI Lab @ IBA. The purpose of this software is to provide businesses an easy to use expert system in which the analysts can easily create and test models and the end-users get predictions for new instances/occurrences. 




Currently the software supports three famous algorithms: classification trees, naive Bayes and neural networks.




The software consists of multiple interfaces for various user-roles. The Anaylst View helps an analyst in model creation and selection. The module allows the analyst to load data from multiple sources and manipulate this data using certain functions like: partitioning data rows, adding/removing columns and filling missing values. The model provides confusion matrix and accuracy report to assist the analyst during model selection stage. The User View helps the enterprise end-user to upload new data instances for which decision is required. The model set by the analyst using the Analyst View is used to predict for the new batch of data.





Updates

  • Osama Khan
    {IBA}Miner, BS '09
    Osama Khan worked on {IBA}Miner as his Final Year Computer Science Project. The project was themed around decision support system for SMBs.

  • Maria Hussain
    {IBA}Miner, BS '09
    Maria Hussain worked on {IBA}Miner as her Final Year Computer Science Project. The project was themed around decision support system for SMBs.

Sitemap