Conventional computing vs Intelligent computing:
Conventional computing is the method of computing technique in which a predefined set of instructions or algorithms are provided to the machine, which are stored in the memory and execute in a sequential way when needed. Whereas in intelligent computing is works on the basis of knowledge base, it uses the knowledge base for reasoning and pattern matching related to a solution of particular problem.
Conventional computing guarantees a consistent and reliable solution for the given problem as the programmer provide the logical steps to reach to a conclusion within in some predefined set of instructions. Whereas intelligent computing doesn’t guarantee a reliable solution for given problem as it uses different type of data to analyze and find out a best solution which may or may not be correct due to its continuous learning process.
Conventional computing can solve only those problems which lies in its operational constraints. Whereas intelligent computing can solve range of problems and also tries to solve new problems as well.
Conventional computing is not always able to process the data which is collected from the external environment whereas an intelligent machine can process the data collected from various sensors.
Conventional computing is not able to improve its solutions whereas intelligent computing improves its solutions because of its learning ability.
Conventional computing map all the possible solutions for a given problem and acts accordingly to achieve its goal whereas intelligent computing is built to learn from past experiences and then take actions to choose the most efficient solution.
Example: A cleaning robot based on conventional computing can clean the floor
But an intelligent robot can clean the floor with maximum efficiency like calculating the shortest path in real time and clean the floor in possible minimum time.
How can we decide if a solution is intelligent:
1. By measuring the accuracy of a solution.
2. An answer will be called an intelligent answer if its based on some reasoning.
3. If the answer is close to the answer given by a human being.
4. The answer is logical and unbiased.
5. Solutions should be efficient and directed towards its goal.
A utility agent is a goal-based agent. Utility agent is based on perception and reasoning.
Lets take an example of cleaning robot which is analyses the environment by using sensors and find out the best cleaning solution but it cannot work efficiently in new condition.
Whereas learning agent consist of learning and performance element which is used in taking the actions in the unknown environment and perform efficiently. Lets take an example of mars rovers it acts in an unknown and new environment by using its learning capabilities which helps it to judge its moves and also learn the new techniques and improve its performance.