Artificial Intelligence Approaches
How its possible that the Artificial Intelligence be put into implementation?
There are various steps to implement Artificial Intelligence. Worldwide speaking, the field of AI is compared between the rule-based techniques and machine learning techniques. The entire world of AI can be divided into these two groups.
A computer system which can achieve AI through the rule-based technique is called as rule-based system and the computer system that achieves AI through a machine learning technique is called a learning system.
What is rule based system?
A rule-based system, lets consider example: production system, expert system which makes use of rules as the knowledge representation. These regulations have been encrypted into the method in the form of if-then-else statements.
Basically the main idea of a rule-based system is to grab the knowledge of a human expert which is particularly the domain and embody it within a computer system and no more or no less. Hence the knowledge is been encoded as the rules.
A rule-based system is like a human born with fixed knowledge. The understanding of that human being will not change over the time. This means that the human being faces a problem for which no rules have been created, then this human gets stuck and so won’t be able to solve the problem. In a sense, the humans will not even even understand the problem.
What is learning system??
In learning systems it has a very ambitious goal. The vision glimpse of AI research, which turns out to be more a desire than a concrete vision, is to implement general AI through the learning capability of these systems. Hence, the desire is that the learning system is in principle unlimited in its ability to simulate intelligence.
It’s said to have adaptive intelligence. The capacity to acquire the knowledge which causes adaptive intelligence, where the adaptive intelligence means that existing knowledge can be changed or it will be discarded, and the new knowledge can be acquired. Hence, these systems been created by the rules on the fly. This is what it is making learning systems so different from the rule-based testing.
The primary goal of a learning system is to do less function, and the system does this by tweaking the weights in such a way that the function is minimised. Therefore the learning just means finding the right way to make less utility function.
What are the problems??
What’s the actual problem with these systems? Although the learning process is deterministic which means it includes statistical and probabilistic methods, which is almost impossible from a practical perspective to extract the model from the internal functioning of a learning system because of its very complexity, and can caused by the gazillions of dynamic parameters.
As a natural this results in the models that have been learned which can’t be interpreted, explained, and also understood well enough by humans. For this cause, we can say that the learning systems are frequently referred to as black boxes which aren’t entirely clear how these systems make their decisions.
What is GDPR??
GDPR is General Data Protection Regulation. It is a data protection law that was implemented by the European Union(EU) in May 2018. Is GDPR a game-changer for AI technologies? It’s not yet clear what it really means and it has remained to be seen whether such a law is been legally enforceable or not and if It’s not even clear if that law is more a right to inform rather than a right to explanation then the impact of GDPR on AI is still heavily debated. It requires the organisations to get clear and explicit consent from the individuals before collecting their personal data.
