Bayesian Networks

What-if analysis from available data Quantum Leap Advisor supports simple what-if analysis based upon informative patterns in data. The Leap Advisor automatically generates hypotheses for root cause analysis and resulting strategies using a Bayesian network. The underlying technology behind Quantum Leap Advisor links advanced optimization technologies with information theory and automatically generated Bayesian network models from pattern rich data to answer “what-if” questions in a unique and flexible manner.

Simulation:

What-if analysis often requires simulations when the future state of the business is too complicated to be extrapolated from historical data. Leap Advisor Simulation is an innovative hybrid simulation platform that combines the strength of agent based modeling and system dynamics to allow users to simulate different scenarios of decision making with flexibility and efficiency.

With agent based modeling paradigms, Quantum Leap Advisor Simulation enables users model emerging complex behaviors by individual entities and the interactions among these entities.

Leap Advisor Simulation has been applied in a wide range of applications, including:

Pandemic Influenza – with US Northern Command who are the designated DOD Global Synchronizer for Pandemic Influenza Response. Pandemic Influenza, Avian Flu and Seasonal Influenza – with Pacific Command in a multi-national exercise (Cobra Gold ’08). Endemic Diseases (Cholera, Pneumonia, Malaria, and Hepatitis A) – with US Marine Forces, Pacific in support of Operation Caring Response. Optimization is used to search for the best response or action among a large space of choices given the limited resources and multiple constraints of the business.

Quantum Leap Advisor Adaptive Optimization® is a patented general-purpose problem-solving system that is especially suited for:

  • Unusually complex, non-traditional optimization problems
  • Situations based on changing or unstable requirements
  • Data-driven scenarios

Quantum Leap Advisor Adaptive Optimization has been applied in a wide range of manufacturing and military applications, including tradeoffs when designing new vehicle concepts, planning, allocation, and scheduling of assembly lines and manufacturing capacity. Easy Data Access Quantum Leap Analyst allows users to access data from any of the sources, including relational databases (e.g., Oracle, MySQL), Excel files and CSV files. Plus, you can view different data tables in a database and merge tables together without using any SQL queries.

Data pre-processing and profiling raw data is highly susceptible to noise, missing values and inconsistency, which can lead to project failures if not properly processed. Quantum Leap Analyst can help users with data pre-processing, including data selection, data cleaning, and data transformation. In addition, the data profiling feature in Leap Analyst enables users to examine various statistics of attributes and understand the quality of the data. Pattern Exploration and Discovery Traditional Business Intelligence (BI) products place a significant burden on the user to be the navigator through charts and tables. This can be time-consuming and unreliable in a complex data environment when there are hundreds of attributes and/or the user might not be knowledgeable enough about potential indicators.

Leap Analyst can automatically discover, organize and visualize key relationships (or meaningful patterns) amongst multiple attributes in a few clicks and deliver insights in minutes. Prediction in a dynamic data environment, a key challenge is to build reusable and transparent predictive models directly from the data. Quantum Leap Analyst allows users to access heterogeneous data sources, develop models for transparent decision making, update models efficiently with new data, and easily tune their models to exhibit desired characteristics.