artificial-intelligence.md

KYT/CAT - Artificial Intelligence

Q.1 About learning reinforcement, what comes hand in hand ?

• Policy and state

Q.2 What should be balanced in reinforcement learning ?

• Exploitation and exploration

Q.3 In robotics, localization comes before planning :

• True

Q.4 What do reinforcement learning methods need ?

• A lot of time

Q.5 Robots can learn new actions all by themselves.

• True

Q.6 What is a knowledge base ?

• A way of encoding information

Q.7 How does the Monte Carlo algorithm work ?

• It uses randomness

Q.8 Roughly estimated over the past ten years, the computing power of smartphones has been multiplied by :

• 1000

Q.9 Evolutionary neural networks use the environment as an input, like reinforcement learning.

• False

Q.10 Who is Elon Musk ?

• All these answers

Q.11 Who is the other co-founder of google, with Larry Page ?

• Sergey Brin

Q.12 What is the goal of Google's BERT ?

• Natural Language Processing

Q.13 Which statement is false about machine language ?

• A word from a machine language can have two meanings

Q.14 When an AI tries to find patterns on its own, it is :

• Unsupervised learning

Q.15 In unsupervised learning, where do the training labels come from ?

• From the environment

Q.16 When babies learn how to walk, it is :

• Unsupervised learning

Q.17 When an AI agent recognizes different properties and creates categories, it is :

• Unsupervised learning

Q.18 While planning, robots need to define :

• Consequences

Q.19 What is a web crawler ?

• A program to gather data

Q.20 Which statement is false ?

• Robots can see dead people

Q.21 Overfitting issues may increase according to :

• Neural network complexity

Q.22 Is precision the only thing that matters for an AI ?

• False

Q.23 When an AI agent tries many different actions, it is :

• Exploring

Q.24 NLP stands for :

• Natural Language Processing

Q.25 In artificial neuron process, which part act as a threshold ?

• Bias

Q.26 What are the 2 main parts of a neural network ?

• Architecture and Weights

Q.27 Deep learning uses old existing techniques ?

• Machine learning

Q.28 What are the major goals of AI ?

• All these goals

Q.29 Chatbots can train themselves by learning from human-to-human conversations ?

• True

Q.30 What are Chinook & Deep Blue ?

• Computers playing games better than humans

Q.31 What is an index ?

• A tool for gathering data

Q.32 AI system could help humans to make decisions because :

• They are good and fast at evaluating many possibilities

Q.33 In which domain does AI get many applpications ?

• All these domains