computational intelligence week 1

What is Computational Intelligence (CI)?

• • A subset of artificial intelligence that focuses on logical reasoning.

• • A set of nature-inspired computational methodologies to address complex real-world problems.

• A programming language used for machine learning.

• A hardware-based approach to computing.

Who first called a system computationally intelligent, and what are its characteristics?

• • Arthur Samuel, focusing on games.

• • Tom M. Mitchell, emphasizing improvement with experience.

• Bezdek in 1994, mentioning pattern-recognition and adaptive mechanisms.

• Simon Haykin, focusing on neural networks.

What are the main techniques used in Computational Intelligence?

• • Fuzzy logic, evolutionary computing, and neural networks.

• Probability theory, linear algebra, and calculus.

• • Hardware computing, cloud storage, and data mining.

• • Software development, web technologies, and database management.

What does Machine Learning (ML) fundamentally rely on to improve performance?

• • Faster computing hardware.

• Larger datasets.

• • Experience, in the form of data.

• • Complex algorithms.

According to Arthur Samuel (1959), how is Machine Learning described?

• • As the development of neural networks.

• As the programming of a computer to behave in a way that involves learning.

• • As the use of fuzzy logic for decision making.

• • As the application of evolutionary computing.

Tom M. Mitchell's definition of Machine Learning involves which components?

• • Experience E, tasks T, and performance measure P.

• • Data D, algorithms A, and results R.

• • Hardware H, software S, and networks N.

• • Programming languages L, databases D, and user interfaces U.

How does machine learning differ from traditional programming?

• ML uses hard-coded rules.

• • ML is data-driven, improving with experience.

• • ML focuses solely on neural networks.

• • ML does not require algorithms.

What is a regression machine learning?

• Predicting continuous label

• predicting discrete label

• working without label

• predicting multiclass label

Which of the following best describes the difference between Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs)?

• • CNNs are primarily used for image classification, while RNNs are used for hardware optimization.

• • CNNs and RNNs are both best suited for image recognition tasks.

• • CNNs are designed to handle spatial hierarchies in data, while RNNs are optimized for sequential data where current inputs are dependent on previous ones.

• • CNNs are used for sequential data processing, while RNNs are used for static image classification.

What is the computational challenge for building a model?

• Big Data

• Big machine

• Big volume

• complicated formula