Artificial intelligence (AI) has immense potential. But it has yet to be leveraged adequately in several fields.
To be valuable, a system has to be intelligent enough to do much more than merely execute its tasks precisely. Unfortunately, many solutions simply use the term artificial intelligence (AI) without deploying any of its algorithms. I remember a friend of mine, who had worked as a CTO in a leading company, being deceived by an ISV (independent software vendor) that sold his company what it claimed was a capital markets prediction software using neural networks. Some solutions claim the deployment of several AI techniques including neural networks, genetic programming and genetic algorithms, but contain nothing! There are several fake solutions available on the World Wide Web, such as some intelligent RSS (really simple syndication) reading solutions and neuro solutions.
But not all AI solutions are bad. There are many good solutions that are used in research fields (and the R&D sectors of industry) such as non-linear regression, forecasting, curve fitting, classification, decision-making, problem optimisation, pattern recognition, and time series analysis. Then there are solutions used for doing business viz., those that help in financial analysis and investment (combining traditional technical analysis with state-of-the-art artificial intelligence technologies), and solutions used as competitive intelligence calculators (CICs). Finally, there exist some solutions employing the back propagation algorithm in AI Microsoft MCTS Training.
How effective is AI?
In theory, AI seems perfect. It has a wide range of applications ranging from network design to embedded solutions. But on the practical side, there are many hurdles. I remember reading a paper on"Using Artificial Intelligence in Communication System Design" by N Shiratori, K Takahashi, K Sugawara and T Kinoshita, which reveals more about the implementation of the theory. There are scientists who believe in a knowledge-based design support system (KDSS) that will assist designers who are inexperienced in communication system design, in creating advanced systems like intelligent networks and large-scale distributed computing systems MCITP Certification.
Large-scale applications are quite similar in their modus operandi to the top layers of the pyramid structure in Game AI. The application of Game AI (which is best thought of as a hierarchy or pyramid) is not just to decision-making. Once employed, it can solve large meshes. Some fellow hackers in our LUG (Linux user group) have even modified some of its algorithms for Password Recovery Master Solutions.
No comments:
Post a Comment