Models: Cortex

Most market models rely on linear assumptions; more or less X implies proportionally more or less Y. But some analytical structures are nonlinear, undefined, or discontinuous.  Neural nets are for addressing such relationships.

Cortex is our proprietary neural net. Like other NNs, Cortex is an adaptive learning model, requiring hands-on implementation. This learning process can be resource intensive.

Unlike algebraic and canonical solutions, NN output gives objective solution (output), but does not provide conceptually recognizable structure.

Neural nets project a cachet in the press, but most market relationships turn out (with a bit of creativity) to work quite well within conventional analytical frameworks.  Cortex is therefore best reserved for special situations.  In complex and theoretically amorphous relationships, Cortex exploration may find otherwise elusive solutions.

Contact Analytix below (or use checklist back on Models page) to enquire about Cortex-based projects.

To enquire about Cortex applications with your own data, you can leave a message back at the Models page, or use the master checklist of Analytix Services.  Or call directly anytime, at (603) 643-6430.

Nothing at this site is offered as advice or recommendation.
All site content is subject to full Disclaimer statement and Terms of Use.
Copyright© 2004. All rights reserved.