Tutorials References Exercises Videos NEW Menu
Paid Courses Website NEW

AI Sciences

Artificial Intelligence is a set of different Sciences

  • Machine Learning (ML)
  • Neural Networks (NN)
  • Deep Learning (DL)
  • Big Data
Weak Machine Learning Neural Networks Big Data Deep Learning Strong

AI Sientists

AI Scientists build software with algorithms that can do tasks that normally require human intelligence.

AI Sientists can be experts in multiple AI disciplines:

  • Applied mathematics
  • Computational statistics
  • Computer Science
  • Machine learning
  • Deep learning

Some AI Scientists also have significant big data experience:

  • Business Inteligence
  • Data Base Design
  • Data Warehouse Design
  • Data Mining
  • SQL Queries
  • SQL Reporting

Weak AI

Weak Artificial Intelligence is limited to specific or narrow areas like most of the AI we have around us today:

  • Search Engines
  • Apple's Siri
  • Microsoft's Cortana
  • Amazon's Alexa
  • IBM's Watson

Weak AI is also called Narrow AI.

Weak AI simulates human cognition in contrast to Strong AI that have human cognition.


Strong AI

Strong Artificial Intelligence is the type of AI that mimics human intelligence.

Strong AI indicates the ability to think, plan, learn, and communicate.

Strong AI is the theoretical next level of AI: True Intelligence.

Strong AI moves towards machines with self-awareness, consciousness, and objective thoughts.

One need not decide if a machine can "think".
One need only decide if a machine can act as intelligently as a human.

Alan Turing


Machine Learning (ML)

Classical programming uses programs to create results:

Traditional Computing

Data + Computer Program = Result

Machine Learning uses results to create programs (algorithms):

Machine Learning

Data + Result = Computer Program

"Machine Learning is a field of study that gives computers the ability to learn without being programmed."

Arthur Samuel (1959)


Neural Networks (NN)

One of the most significant discoveries in history is the power of Neural Networks (NN).

In Neural Networks, many layers of data called Neurons are added together or stacked on top of each other to compute new levels of data.

Commonly used short names:

  • DNN Deep Neural Network
  • CNN Convolutional Neural Network
  • RNN Recurrent Neural Network

Deep Learning (DL)

Deep Learning are algorithms that use Neural Networks to extract higher-level data.

Each successive layer uses the preceding layer as input.

For instance, optical reading uses low layers to identify edges, and higher layers to identify letters.

Deep Learning has two phases:

1. Training: Input data are used to calculate the parameters of the model.

2. Inference: The "trained" model outputs data from any given input.


The Deep Learning Revolution

The deep learning revolution is here!

The deep learning revolution started around 2010. Since then, Deep Learning has been used to solve many "unsolvable" problems.


Examples

Convolutional Neural Networks (CNNs)

Deep CNNs such as ResNeta and Inception have reduced the error rate in the ImageNet classification from 25% in 2011 to 5% in 2017.

ImageNet is an image database organized according to the WordNet hierarchy, in which each node of the hierarchy contains hundreds and thousands of images. ImageNet is a useful resource for researchers, educators, students and everyone else with a passion for pictures.

WordNet is a lexical database of semantic relations between words in 200+ languages. It is organized as a combination of a dictionary and thesaurus, linking words together into semantic relations using synonyms, hyponyms, and meronyms.

Recurrent Neural Networks (RNNs)

RNNs are helping create music scores and novel instrument sounds:
https://magenta.tensorflow.org/demos.


History of AI

1950Alan Turing publishes "Computing Machinery and Intelligence"
1956AI first mentioned by John McCarthy in an academic conference
1957First programming language for numeric and scientific computing (FORTRAN)
1958First AI programming language (Lisp)
1959Arthur Samuel used the term "Machine Learning"
1961First industrial robot (Unimate) on the assembly line at General Motors.
1965ELIZA by Joseph Weizenbaum was the first interactive program that could communicate on any topic
1972First logic programming language (PROLOG)
1997Deep Blue (IBM) beats the world champion in chess
2002The first robot cleaner (Roomba)
2005Self-driving car (STANLEY) wins DARPA
2008Breakthrough in speech recognition (Google)
2011A neural network wins over humans in traffic sign recognition (99.46% vs 99.22%)
2011Apple Siri
2011Watson (IBM) wins Jeopardy!
2014Amazon Alexa
2014Microsoft Cortana
2014Self-driving car (Google) passes a state driving test
2015Google AlphaGo defeated various human champions in the board game Go
2016The human robot Sofia by Hanson Robotics
Sopia