TU Vienna revolutionizes KI: New method makes uncertainty measurable!
TU Vienna revolutionizes KI: New method makes uncertainty measurable!
What's new from research at the Technical University of Vienna? A committed research team has developed an innovative mathematical method to calculate uncertainty in neuronal networks. This method could revolutionize the way we make decisions based on artificial intelligence. In the world of AI, where there is often a lack of transparency, this is a real ray of hope.
The challenge for classic neuronal networks is obvious: they often deliver the same results when they are identical, but small changes can cause unexpected results. So far, these systems do not provide any information about the security of their decisions. According to the researchers of the Vienna University of Technology, it is time for a rethink. The new procedure based on a geometric approach is broken down into small sections. For each of these areas, the possible expenses can be calculated exactly, which enables effortless quantification of the error probabilities.
mathematical and technical sophistication
The special activation functions that are used in relu networks are particularly exciting. These are described by two approximation networks, which simplifies the arithmetic operations enormously. In tests on data records such as "Iris", "Wine", "Diabetes" and "Banana", the researchers have shown that their method offers more precise mathematical barriers for network expenses than previous approaches. It turned out that the method works particularly efficiently in low -dimensional input dimensions and also does not lure you to accuracy even in more complex challenges - but at the expense of a higher computing effort.A noticeable advantage of this method is your focus on smaller, controllable neural networks. However, more complex AI systems, such as chatt, cannot be analyzed. This finding opens up new perspectives, especially in security -critical areas such as medical technology, robotics, the development of autonomous vehicles and the financial risk assessment. The general goal of the developers is to improve decision making by better understanding of the probabilities of possible results.
practical applications of research
The potential possible uses of this new method are diverse and could have far -reaching effects on many areas of life. This technology could soon be used in fields, where security and accuracy are very priority. Whether in the healthcare system or in the financial industry - the opportunity to calculate uncertainties and to better justify decisions is very important.
Research work at the Technical University of Vienna not only opens up new opportunities for artificial intelligence, but also sets an important impulse for further development in this exciting and future -oriented area. Researchers and companies are now called to take up this new methodology and to continue thinking in their own developments. The future of the AI could thus become clearer and more predictable.
For more information about the new method and its effects on research and industry, read the detailed reports at and TU Vienna .
| Details | |
|---|---|
| Ort | Wieden, Österreich |
| Quellen | |
Kommentare (0)