Lone Star Analysis, a Dallas-based specialist in leading-edge predictive and prescriptive analytics solutions, uses glass-box solutions with unparalleled transparency. These solutions provide clients clear-cut, explainable insights from Lone Star's services and software.
The alternative, or black-box, solutions are fueled by artificial intelligence (AI) using unexplainable means to reach outcomes. Most AI is based on complex neural networks which are impossible to explain. Decision makers find these results hard to accept. About 60 percent of 5,000 executives in an IBM survey expressed concern about black-box solutions.
“According to our own research, black-box solutions are always harder for people to accept, and this sentiment is growing stronger,” said Steve Roemerman, CEO, Lone Star Analysis. “Our glass-box solutions provide clarity into the data, logic and math used for analysis. This results in reliable, quantifiable and valuable recommendations for clients.”
New privacy laws restrict data uses and make brute-force, black-box AI problematic. These schemes require large training data sets, and the new laws hinder the use of consumers' data. This adds both legal and reputation risks. Meanwhile, Lone Star can help clients make smarter decisions faster by using smaller data sets with much less risk.
The combination of algorithm transparency, with small, low risk data sets is central to Lone Star's understandable methods.
Lone Star's definition of transparency includes explainable analytics. This enables conclusions to be proven and reproduced. With Lone Star's glass-box solutions, data and reasoning are accessible and open to scrutiny. This provides a more reliable and trustworthy recommendation than any black-box scheme.