|
Success Stories Read about just a few of our customer's
success stories.
Here are links to comments about Discipulus, the scientific and engineering data mining
and modeling tool based on AIMLearning Technology:
"Discipulus is one of the strongest predictive modeling packages in history. It is blazing fast, it evolves accurate, compelling and deployable models, and it is a stunning example of ingenuity and tradecraft, and certainly the finest example of machine based Linear Genetic Programming ever designed. For a few hundred dollars, you will be buying one of the greatest predictive modeling packages ever built. I have one word to describe the predictive capabilities of Discipulus…. WOW !!....I paid about $500 for the Professional Version, and now that I see the capabilities of Discipulus, I would have gladly paid $10,000 for this software package. Discipulus is a VERY strong predictive modeling package, and I distinctly recall the first time I used the software, watching it evolve a simple mathematical answer, to a complex modeling problem, my jaw dropped, and I said to myself “Toto, we are not in Kansas anymore...."
Brian Watt, CRM
Chief Risk Officer, Chief Financial Officer, GECC, Inc
"Discipulus is an outstanding product. It recently solved an
extremely difficult process control problem for us that had stumped our best neural
network researchers.
"Other machine learning systems would take well over a year
in CPU time on problems Discipulus comfortably solves in a week. It produces a simulation
model of the process, which is designed for direct incorporation into scientific data
models and process controllers.
"Thank you, RML Technologies."
Larry M. Deschaine, PE, Science Applications International Corp.
"Discipulus is a fast and robust tool for
transforming data into equations. I have personally used the tool since
1998 and is my tool of choice for regression and classification in noisy
data. With it, I have been able to solve problems that other methods
(neural networks, linear regression, decision trees, SVM's etc) were
unable to solve during well conducted studies.
Discipulus uses a combination of machine code based genetic algorithms
and evolutionary strategies to uncover the mathematical program that
represents the functional information content of the data. It is
lightning fast compared to other evolutionary programming approaches and
quite immune to overfitting.
"I have successfully used Discipulus to detect buried bombs from digital
signals, predict soil properties from scientific explorations, predict
leachability of radionucleotide(s), predict processes in nuclear power
plants, and even some interesting features of satellite imagery, etc. I
even tried to deceptively trick it into modeling noise but was unable to
do so."
Model &
Mine Review, 2004
"We are very pleased with this product. It's fast and the
user interface is good. . . In the past year, Discipulus has been the key to fast
deployment of several important in-house modeling solutions."
Emerging Technologies Director, Fortune 100 Company
"I am quite happy with the program. It is well documented
with a comprehensive user/reference manual and a nice set of tutorials. After quickly
reading the user manual, it took me less than an hour to . . . complete my first run.
"As their motto says, 'Speed Matters.' Their system is quite
fast. I also have to give very high marks to their customer support.
Dan Yagusic, Microchip Defect Analysis Engineer
"At last, a commercial, industrial strength system for
genetic programming to solve practical problems."
Dr. James Foster. Machine Learning Consultant and Prof. Computer
Science, University of Idaho
"Until you have access to [modeling] tools this fast, you
really cannot appreciate how much of a difference it makes."
EvoNews, Winter 1999.
|