BioComp Systems

The name BioComp comes from Biologically based Computing algorithms.

Using the Power and Elegance of Nature

Natural algorithms, such as neural computing, genetic algorithms, swarm optimization and adaptation, have had billions of years' success in creating diversity while seeking ever-better solutions. We draw upon these powerful algorithms combined with modern computer science to deliver you technologies and solutions that have direct beneficial impact on your bottom-line.

Neural Computing

Neural networks, and our advance upon that; Meshes, combined with proper data handling and repeated validation, have demonstrated superior abiity to perform linear / non-linear regression, probability estimation and clustering analysis. They are the best technology for modeling, prediction and through advanced non-linear seach technology, they are capable of model-based optimal control. Such models are analyzed by our technologies to yield "sensitivity analysis", an expression of what variables drive results to what degree, invaluable for seeing, understanding and enhancing business, product and process performance and diagnostics.

Genetic Algorithms

Genetic Algorithms (GAs) are an outstanding approach for rapidly searching through billions of possible alternatives to create the best solution. We created the 1st commercial GA library for Microsoft Windows back in the 1980's and use the technology extensively in our products to this day. Nearly every product shipped has a GA on-board. We use them to:

  • Search for the best combination of inputs to neural-like models, your performance's top causal drivers, while at the same time searching for the best model type and internal structure of neural-like models to predict performance, forecast demand, estimate future probabilities of events, likelihoods of machine failures and many other applications.
  • Compress production schedules by seeking the best combinations of task sequences that meet priorities and reduce overall work durations.
  • Find the best actions to take to achieve stated objectives, such as determining setpoints in model-predictive control.

Swarm Adaptation

Since the world is always changing, you need to adapt to survive. Swarm Adaptation algorithms, such as the behavior of a flock of birds, is an excellent choice to cover a search space while always moving towards enhanced performance. Different than "Swarm Optimization", seeking *THE* best solution, Swarm Adaptation uses an inter-communicating group of solutions that avoid getting trapped in sub-optimal solutions, having the advantage of many simultaneous perspectives within fruitful regions of performance.