SlideShare a Scribd company logo
1 of 17
Download to read offline
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Andrew	
  Jeavons	
  
Mass	
  Cogni+on	
  
	
  
	
  
18	
  August	
  2016	
  
An	
  Introduc+on	
  to	
  
Neural	
  Networks	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
The	
  Beginning	
  
•  “A	
  Logical	
  Calculus	
  of	
  the	
  ideas	
  immanent	
  in	
  
Nervous	
  Ac2vity”	
  
–  McCulloch	
  and	
  PiDs,	
  1943	
  
•  Organiza2on	
  of	
  neurons	
  and	
  logic.	
  
–  Mimicking	
  organiza2on	
  of	
  neurons	
  in	
  brain	
  
•  Beginning	
  of	
  “neural	
  networks”	
  approach	
  
–  But	
  not	
  a	
  simple	
  path.	
  
	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Neuron:	
  All	
  or	
  Nothing	
  
What	
  comes	
  out	
  is	
  not	
  necessarily	
  
what	
  goes	
  in.	
  
	
  
“Transfer	
  Func2on”	
  –	
  controls	
  what	
  the	
  	
  
output	
  is.	
  
	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Ar+ficial	
  Neuron:	
  All	
  or	
  Nothing	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Ar+ficial	
  Neuron	
  
SoTware	
  “copy”	
  of	
  the	
  structure	
  of	
  a	
  physical	
  neuron.	
  
	
  
Much	
  simplified,	
  many	
  aspects	
  of	
  neuron	
  func2on	
  were/are	
  hard	
  
to	
  simulate.	
  
	
  
Arranged	
  into	
  more	
  complex	
  mul2	
  layer	
  structures.	
  
	
  
	
  
	
  
	
  
	
  
	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
ANN:	
  Backpropoga+on	
  Network	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Types	
  of	
  Training:	
  Supervised	
  
Supervised:	
  
	
  Weights	
  adjusted	
  itera2vely	
  to	
  bring	
  down	
  the	
  difference	
  between	
  the	
  	
  
	
  output	
  generated	
  by	
  the	
  ANN	
  by	
  the	
  training	
  data	
  set	
  and	
  the	
  test	
  value.	
  	
  
	
  
For	
  supervised	
  learning	
  divide	
  training	
  data	
  set	
  into	
  train	
  and	
  test	
  segments.	
  
	
  
ANN	
  learns	
  on	
  the	
  training	
  data	
  set	
  and	
  tests	
  this	
  training	
  on	
  the	
  test	
  dataset.	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Types	
  of	
  Training:	
  Unsupervised	
  
Usually	
  different	
  network	
  structure.	
  For	
  example	
  a	
  Kohonen	
  Self	
  
Organizing	
  Map:	
  
All	
  inputs	
  connect	
  to	
  ALL	
  outputs.	
  Can	
  have	
  hidden	
  layers.	
  Acts	
  as	
  a	
  classifier,	
  
produces	
  a	
  map	
  (similar	
  to	
  segmenta2on)	
  of	
  the	
  input	
  data.	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Types	
  of	
  Training:	
  Unsupervised	
  
Example	
  of	
  Kohonen	
  Map:	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Types	
  of	
  Training:	
  Unsupervised	
  
Select	
  an	
  input	
  at	
  random	
  
	
  
Find	
  which	
  ar2ficial	
  neuron	
  is	
  most	
  similar	
  
based	
  on	
  weight	
  value.	
  
	
  
Adjust	
  values	
  of	
  “winner”	
  neuron	
  to	
  be	
  close	
  to	
  	
  
Input.	
  
	
  
Keep	
  doing	
  this.	
  
	
  
Then	
  you	
  can	
  map	
  a	
  new	
  input	
  “vector”	
  and	
  	
  
get	
  classifica2on.	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
•  Lots	
  of	
  arithme2c.	
  
–  Video	
  games	
  GPU’s	
  have	
  helped	
  alleviate	
  this	
  problem.	
  
•  Complex	
  
–  As	
  bigger	
  memory	
  capacity	
  became	
  available	
  bigger	
  
networks	
  feasible.	
  
•  Coding	
  of	
  input.	
  
•  Architecture	
  selec2on.	
  
–  Thousands	
  of	
  architectures	
  available.	
  
	
  
ANN:	
  Challenges	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
ANN:	
  New	
  and	
  Interes+ng	
  
•  “Genera2ve”	
  Networks.	
  
–  Produce	
  new	
  output	
  rather	
  than	
  classifica2on	
  and	
  
predic2on.	
  
•  Large	
  scale	
  networks	
  for	
  text	
  analysis.	
  
–  Google	
  word2vec	
  and	
  doc2vec.	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
ANN:	
  LSTM	
  Recurrent	
  ANN	
  	
  
•  Long	
  Term	
  Short	
  Term	
  Memory	
  ANN.	
  
–  Used	
  to	
  generate	
  a	
  short	
  movie.	
  
–  Trained	
  on	
  scripts	
  of	
  Scifi	
  movies/TV	
  shows	
  on	
  interweb.	
  
–  (hDp://arstechnica.com/the-­‐mul2verse/2016/06/an-­‐ai-­‐
wrote-­‐this-­‐movie-­‐and-­‐its-­‐strangely-­‐moving/)	
  
–  AI	
  came	
  up	
  with	
  its	
  own	
  name,	
  Benjamin,	
  in	
  an	
  interview.	
  
–  Biologically	
  Plausible	
  ANN,	
  uses	
  cogni2ve	
  psychology	
  
constructs.	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
ANN:	
  Doc2vec	
  
•  Word2vec/doc2vec	
  
–  Take	
  a	
  word	
  or	
  document	
  
–  Create	
  a	
  “vector”	
  (list	
  of	
  numbers)	
  300+	
  long.	
  
–  All	
  vectors	
  are	
  fixed	
  length	
  
–  Vectors	
  can	
  be	
  compared	
  
•  word2vec	
  can	
  show	
  “odd	
  man	
  out”	
  
–  “cat,	
  dog,	
  dolphin,	
  chair”	
  
–  Uses	
  “distance”	
  between	
  vectors	
  to	
  decide	
  odd	
  man	
  out	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
ANN:	
  Doc2vec	
  
•  Doc2vec	
  does	
  the	
  same	
  thing	
  for	
  documents.	
  
–  Converts	
  documents	
  to	
  fixed	
  length	
  vectors.	
  
–  Can	
  calculate	
  the	
  similarity	
  of	
  documents.	
  
–  Can	
  be	
  interrogated	
  as	
  in	
  “which	
  color	
  is	
  best”	
  
•  Prints	
  out	
  most	
  similar	
  documents	
  to	
  input	
  .	
  
•  Input	
  phrase	
  does	
  not	
  have	
  to	
  be	
  in	
  text	
  corpus	
  trained	
  with.	
  
•  Encodes	
  seman2cs.	
  
•  Poten2al	
  for	
  mapping	
  and	
  exploring	
  text	
  
•  Computa2onally	
  intensive	
  !	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Thank	
  You!	
  
	
  
	
  
	
  
www.masscogni2on.com	
  
	
  
apj@masscogni2on.com	
  
An	
  Introduc+on	
  to	
  Neural	
  Networks	
  –	
  Agility,	
  Automa+on	
  &	
  AI	
  
Andrew	
  Jeavons,	
  Mass	
  Cogni2on,	
  2016	
  
Q	
  &	
  A	
  
Ray	
  Poynter	
  
The	
  Future	
  Place	
  
Andrew	
  Jeavons	
  
Mass	
  Cogni2on	
  

More Related Content

Similar to Andrew Jeavons AAA 2016

Neural networks and deep learning
Neural networks and deep learningNeural networks and deep learning
Neural networks and deep learningRADO7900
 
Predictive Metabonomics
Predictive MetabonomicsPredictive Metabonomics
Predictive MetabonomicsMarilyn Arceo
 
Artificial Neural Networks ppt.pptx for final sem cse
Artificial Neural Networks  ppt.pptx for final sem cseArtificial Neural Networks  ppt.pptx for final sem cse
Artificial Neural Networks ppt.pptx for final sem cseNaveenBhajantri1
 
Neural Networks Ver1
Neural  Networks  Ver1Neural  Networks  Ver1
Neural Networks Ver1ncct
 
Deep learning tutorial 9/2019
Deep learning tutorial 9/2019Deep learning tutorial 9/2019
Deep learning tutorial 9/2019Amr Rashed
 
Deep Learning Tutorial
Deep Learning TutorialDeep Learning Tutorial
Deep Learning TutorialAmr Rashed
 
Corinna Cortes, Head of Research, Google, at MLconf NYC 2017
Corinna Cortes, Head of Research, Google, at MLconf NYC 2017Corinna Cortes, Head of Research, Google, at MLconf NYC 2017
Corinna Cortes, Head of Research, Google, at MLconf NYC 2017MLconf
 
The Introduction to Neural Networks.ppt
The Introduction to Neural Networks.pptThe Introduction to Neural Networks.ppt
The Introduction to Neural Networks.pptmoh2020
 
Artificial neural network by arpit_sharma
Artificial neural network by arpit_sharmaArtificial neural network by arpit_sharma
Artificial neural network by arpit_sharmaEr. Arpit Sharma
 
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in RUnderstanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in RManish Saraswat
 
Automatic Attendace using convolutional neural network Face Recognition
Automatic Attendace using convolutional neural network Face RecognitionAutomatic Attendace using convolutional neural network Face Recognition
Automatic Attendace using convolutional neural network Face Recognitionvatsal199567
 
Artificial neural networks
Artificial neural networks Artificial neural networks
Artificial neural networks ShwethaShreeS
 
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...Simplilearn
 
Neural networks are parallel computing devices.docx.pdf
Neural networks are parallel computing devices.docx.pdfNeural networks are parallel computing devices.docx.pdf
Neural networks are parallel computing devices.docx.pdfneelamsanjeevkumar
 

Similar to Andrew Jeavons AAA 2016 (20)

Neural networks and deep learning
Neural networks and deep learningNeural networks and deep learning
Neural networks and deep learning
 
Predictive Metabonomics
Predictive MetabonomicsPredictive Metabonomics
Predictive Metabonomics
 
ANN - UNIT 1.pptx
ANN - UNIT 1.pptxANN - UNIT 1.pptx
ANN - UNIT 1.pptx
 
Artificial Neural Networks ppt.pptx for final sem cse
Artificial Neural Networks  ppt.pptx for final sem cseArtificial Neural Networks  ppt.pptx for final sem cse
Artificial Neural Networks ppt.pptx for final sem cse
 
Neural Networks
Neural NetworksNeural Networks
Neural Networks
 
Neural_Network
Neural_NetworkNeural_Network
Neural_Network
 
Neural Networks Ver1
Neural  Networks  Ver1Neural  Networks  Ver1
Neural Networks Ver1
 
Deep learning tutorial 9/2019
Deep learning tutorial 9/2019Deep learning tutorial 9/2019
Deep learning tutorial 9/2019
 
Deep Learning Tutorial
Deep Learning TutorialDeep Learning Tutorial
Deep Learning Tutorial
 
Corinna Cortes, Head of Research, Google, at MLconf NYC 2017
Corinna Cortes, Head of Research, Google, at MLconf NYC 2017Corinna Cortes, Head of Research, Google, at MLconf NYC 2017
Corinna Cortes, Head of Research, Google, at MLconf NYC 2017
 
The Introduction to Neural Networks.ppt
The Introduction to Neural Networks.pptThe Introduction to Neural Networks.ppt
The Introduction to Neural Networks.ppt
 
Artificial Neural Network
Artificial Neural NetworkArtificial Neural Network
Artificial Neural Network
 
Artificial neural network by arpit_sharma
Artificial neural network by arpit_sharmaArtificial neural network by arpit_sharma
Artificial neural network by arpit_sharma
 
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in RUnderstanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
 
Automatic Attendace using convolutional neural network Face Recognition
Automatic Attendace using convolutional neural network Face RecognitionAutomatic Attendace using convolutional neural network Face Recognition
Automatic Attendace using convolutional neural network Face Recognition
 
Neural Networks
Neural NetworksNeural Networks
Neural Networks
 
Artificial neural networks
Artificial neural networks Artificial neural networks
Artificial neural networks
 
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
What Is Deep Learning? | Introduction to Deep Learning | Deep Learning Tutori...
 
Artificial neural network
Artificial neural networkArtificial neural network
Artificial neural network
 
Neural networks are parallel computing devices.docx.pdf
Neural networks are parallel computing devices.docx.pdfNeural networks are parallel computing devices.docx.pdf
Neural networks are parallel computing devices.docx.pdf
 

More from Ray Poynter

The State of AI in Insights and Research 2024: Results and Findings
The State of AI in Insights and Research 2024: Results and FindingsThe State of AI in Insights and Research 2024: Results and Findings
The State of AI in Insights and Research 2024: Results and FindingsRay Poynter
 
ResearchWiseAI - an artificial intelligence driven research data analysis tool
ResearchWiseAI - an artificial intelligence driven research data analysis toolResearchWiseAI - an artificial intelligence driven research data analysis tool
ResearchWiseAI - an artificial intelligence driven research data analysis toolRay Poynter
 
AI-powered interviewing: Best practices from Yasna
AI-powered interviewing: Best practices from YasnaAI-powered interviewing: Best practices from Yasna
AI-powered interviewing: Best practices from YasnaRay Poynter
 
Artificial Intelligence and Qual: The Story So Far
Artificial Intelligence and Qual: The Story So FarArtificial Intelligence and Qual: The Story So Far
Artificial Intelligence and Qual: The Story So FarRay Poynter
 
State of Research Insights in Q1, 2024 from NewMR
State of Research Insights in Q1, 2024 from NewMRState of Research Insights in Q1, 2024 from NewMR
State of Research Insights in Q1, 2024 from NewMRRay Poynter
 
Sudden Death of Beliefs
Sudden Death of BeliefsSudden Death of Beliefs
Sudden Death of BeliefsRay Poynter
 
Uncovering Consumers’ Hidden Narratives
Uncovering Consumers’ Hidden NarrativesUncovering Consumers’ Hidden Narratives
Uncovering Consumers’ Hidden NarrativesRay Poynter
 
Narrative Exploration of New Categories at Mondelēz
Narrative Exploration of New Categories at MondelēzNarrative Exploration of New Categories at Mondelēz
Narrative Exploration of New Categories at MondelēzRay Poynter
 
The Future in Focus
The Future in FocusThe Future in Focus
The Future in FocusRay Poynter
 
The Future in Focus
The Future in FocusThe Future in Focus
The Future in FocusRay Poynter
 
The State of Insights – September 2023
The State of Insights – September 2023The State of Insights – September 2023
The State of Insights – September 2023Ray Poynter
 
Research Thinking in the age of AI
Research Thinking in the age of AIResearch Thinking in the age of AI
Research Thinking in the age of AIRay Poynter
 
How might AI impact Research and Insights over the next two years?
How might AI impact Research and Insights over the next two years?How might AI impact Research and Insights over the next two years?
How might AI impact Research and Insights over the next two years?Ray Poynter
 
From Words to Wisdom: Unleashing the Potential of Language Models for Human-C...
From Words to Wisdom: Unleashing the Potential of Language Models for Human-C...From Words to Wisdom: Unleashing the Potential of Language Models for Human-C...
From Words to Wisdom: Unleashing the Potential of Language Models for Human-C...Ray Poynter
 
ChatGPT for Social Media Listening: practical application with YouScan’s Insi...
ChatGPT for Social Media Listening: practical application with YouScan’s Insi...ChatGPT for Social Media Listening: practical application with YouScan’s Insi...
ChatGPT for Social Media Listening: practical application with YouScan’s Insi...Ray Poynter
 
Using Generative AI to Assess the Quality of Open-Ended Responses in Surveys
Using Generative AI to Assess the Quality of Open-Ended Responses in SurveysUsing Generative AI to Assess the Quality of Open-Ended Responses in Surveys
Using Generative AI to Assess the Quality of Open-Ended Responses in SurveysRay Poynter
 
Exploring the future of verbatim coding with ChatGPT
Exploring the future of verbatim coding with ChatGPTExploring the future of verbatim coding with ChatGPT
Exploring the future of verbatim coding with ChatGPTRay Poynter
 
Using Generative AI to bring Qualitative Capabilities to Quantitative Surveys
Using Generative AI to bring Qualitative Capabilities to Quantitative SurveysUsing Generative AI to bring Qualitative Capabilities to Quantitative Surveys
Using Generative AI to bring Qualitative Capabilities to Quantitative SurveysRay Poynter
 
How AI / ChatGPT Drives Business Growth
How AI / ChatGPT Drives Business GrowthHow AI / ChatGPT Drives Business Growth
How AI / ChatGPT Drives Business GrowthRay Poynter
 
Tech for tech’s sake? Learnings from experiments with AI in consumer research
Tech for tech’s sake? Learnings from experiments with AI in consumer researchTech for tech’s sake? Learnings from experiments with AI in consumer research
Tech for tech’s sake? Learnings from experiments with AI in consumer researchRay Poynter
 

More from Ray Poynter (20)

The State of AI in Insights and Research 2024: Results and Findings
The State of AI in Insights and Research 2024: Results and FindingsThe State of AI in Insights and Research 2024: Results and Findings
The State of AI in Insights and Research 2024: Results and Findings
 
ResearchWiseAI - an artificial intelligence driven research data analysis tool
ResearchWiseAI - an artificial intelligence driven research data analysis toolResearchWiseAI - an artificial intelligence driven research data analysis tool
ResearchWiseAI - an artificial intelligence driven research data analysis tool
 
AI-powered interviewing: Best practices from Yasna
AI-powered interviewing: Best practices from YasnaAI-powered interviewing: Best practices from Yasna
AI-powered interviewing: Best practices from Yasna
 
Artificial Intelligence and Qual: The Story So Far
Artificial Intelligence and Qual: The Story So FarArtificial Intelligence and Qual: The Story So Far
Artificial Intelligence and Qual: The Story So Far
 
State of Research Insights in Q1, 2024 from NewMR
State of Research Insights in Q1, 2024 from NewMRState of Research Insights in Q1, 2024 from NewMR
State of Research Insights in Q1, 2024 from NewMR
 
Sudden Death of Beliefs
Sudden Death of BeliefsSudden Death of Beliefs
Sudden Death of Beliefs
 
Uncovering Consumers’ Hidden Narratives
Uncovering Consumers’ Hidden NarrativesUncovering Consumers’ Hidden Narratives
Uncovering Consumers’ Hidden Narratives
 
Narrative Exploration of New Categories at Mondelēz
Narrative Exploration of New Categories at MondelēzNarrative Exploration of New Categories at Mondelēz
Narrative Exploration of New Categories at Mondelēz
 
The Future in Focus
The Future in FocusThe Future in Focus
The Future in Focus
 
The Future in Focus
The Future in FocusThe Future in Focus
The Future in Focus
 
The State of Insights – September 2023
The State of Insights – September 2023The State of Insights – September 2023
The State of Insights – September 2023
 
Research Thinking in the age of AI
Research Thinking in the age of AIResearch Thinking in the age of AI
Research Thinking in the age of AI
 
How might AI impact Research and Insights over the next two years?
How might AI impact Research and Insights over the next two years?How might AI impact Research and Insights over the next two years?
How might AI impact Research and Insights over the next two years?
 
From Words to Wisdom: Unleashing the Potential of Language Models for Human-C...
From Words to Wisdom: Unleashing the Potential of Language Models for Human-C...From Words to Wisdom: Unleashing the Potential of Language Models for Human-C...
From Words to Wisdom: Unleashing the Potential of Language Models for Human-C...
 
ChatGPT for Social Media Listening: practical application with YouScan’s Insi...
ChatGPT for Social Media Listening: practical application with YouScan’s Insi...ChatGPT for Social Media Listening: practical application with YouScan’s Insi...
ChatGPT for Social Media Listening: practical application with YouScan’s Insi...
 
Using Generative AI to Assess the Quality of Open-Ended Responses in Surveys
Using Generative AI to Assess the Quality of Open-Ended Responses in SurveysUsing Generative AI to Assess the Quality of Open-Ended Responses in Surveys
Using Generative AI to Assess the Quality of Open-Ended Responses in Surveys
 
Exploring the future of verbatim coding with ChatGPT
Exploring the future of verbatim coding with ChatGPTExploring the future of verbatim coding with ChatGPT
Exploring the future of verbatim coding with ChatGPT
 
Using Generative AI to bring Qualitative Capabilities to Quantitative Surveys
Using Generative AI to bring Qualitative Capabilities to Quantitative SurveysUsing Generative AI to bring Qualitative Capabilities to Quantitative Surveys
Using Generative AI to bring Qualitative Capabilities to Quantitative Surveys
 
How AI / ChatGPT Drives Business Growth
How AI / ChatGPT Drives Business GrowthHow AI / ChatGPT Drives Business Growth
How AI / ChatGPT Drives Business Growth
 
Tech for tech’s sake? Learnings from experiments with AI in consumer research
Tech for tech’s sake? Learnings from experiments with AI in consumer researchTech for tech’s sake? Learnings from experiments with AI in consumer research
Tech for tech’s sake? Learnings from experiments with AI in consumer research
 

Recently uploaded

SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxiammrhaywood
 
Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3JemimahLaneBuaron
 
A Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformA Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformChameera Dedduwage
 
Web & Social Media Analytics Previous Year Question Paper.pdf
Web & Social Media Analytics Previous Year Question Paper.pdfWeb & Social Media Analytics Previous Year Question Paper.pdf
Web & Social Media Analytics Previous Year Question Paper.pdfJayanti Pande
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdfSoniaTolstoy
 
Holdier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfHoldier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfagholdier
 
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in DelhiRussian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhikauryashika82
 
Key note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdfKey note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdfAdmir Softic
 
social pharmacy d-pharm 1st year by Pragati K. Mahajan
social pharmacy d-pharm 1st year by Pragati K. Mahajansocial pharmacy d-pharm 1st year by Pragati K. Mahajan
social pharmacy d-pharm 1st year by Pragati K. Mahajanpragatimahajan3
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAssociation for Project Management
 
fourth grading exam for kindergarten in writing
fourth grading exam for kindergarten in writingfourth grading exam for kindergarten in writing
fourth grading exam for kindergarten in writingTeacherCyreneCayanan
 
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...Sapna Thakur
 
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Sapana Sha
 
Grant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingGrant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingTechSoup
 
Arihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfArihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfchloefrazer622
 
9548086042 for call girls in Indira Nagar with room service
9548086042  for call girls in Indira Nagar  with room service9548086042  for call girls in Indira Nagar  with room service
9548086042 for call girls in Indira Nagar with room servicediscovermytutordmt
 
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...fonyou31
 
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...christianmathematics
 

Recently uploaded (20)

SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
 
Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3
 
A Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformA Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy Reform
 
Web & Social Media Analytics Previous Year Question Paper.pdf
Web & Social Media Analytics Previous Year Question Paper.pdfWeb & Social Media Analytics Previous Year Question Paper.pdf
Web & Social Media Analytics Previous Year Question Paper.pdf
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
 
Holdier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfHoldier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdf
 
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in DelhiRussian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
 
Key note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdfKey note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdf
 
social pharmacy d-pharm 1st year by Pragati K. Mahajan
social pharmacy d-pharm 1st year by Pragati K. Mahajansocial pharmacy d-pharm 1st year by Pragati K. Mahajan
social pharmacy d-pharm 1st year by Pragati K. Mahajan
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across Sectors
 
fourth grading exam for kindergarten in writing
fourth grading exam for kindergarten in writingfourth grading exam for kindergarten in writing
fourth grading exam for kindergarten in writing
 
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptxINDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
 
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
 
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
 
Grant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingGrant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy Consulting
 
Arihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfArihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdf
 
9548086042 for call girls in Indira Nagar with room service
9548086042  for call girls in Indira Nagar  with room service9548086042  for call girls in Indira Nagar  with room service
9548086042 for call girls in Indira Nagar with room service
 
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
 
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...
 
Mattingly "AI & Prompt Design: The Basics of Prompt Design"
Mattingly "AI & Prompt Design: The Basics of Prompt Design"Mattingly "AI & Prompt Design: The Basics of Prompt Design"
Mattingly "AI & Prompt Design: The Basics of Prompt Design"
 

Andrew Jeavons AAA 2016

  • 1. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Andrew  Jeavons   Mass  Cogni+on       18  August  2016   An  Introduc+on  to   Neural  Networks  
  • 2. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   The  Beginning   •  “A  Logical  Calculus  of  the  ideas  immanent  in   Nervous  Ac2vity”   –  McCulloch  and  PiDs,  1943   •  Organiza2on  of  neurons  and  logic.   –  Mimicking  organiza2on  of  neurons  in  brain   •  Beginning  of  “neural  networks”  approach   –  But  not  a  simple  path.    
  • 3. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Neuron:  All  or  Nothing   What  comes  out  is  not  necessarily   what  goes  in.     “Transfer  Func2on”  –  controls  what  the     output  is.    
  • 4. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Ar+ficial  Neuron:  All  or  Nothing  
  • 5. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Ar+ficial  Neuron   SoTware  “copy”  of  the  structure  of  a  physical  neuron.     Much  simplified,  many  aspects  of  neuron  func2on  were/are  hard   to  simulate.     Arranged  into  more  complex  mul2  layer  structures.              
  • 6. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   ANN:  Backpropoga+on  Network  
  • 7. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Types  of  Training:  Supervised   Supervised:    Weights  adjusted  itera2vely  to  bring  down  the  difference  between  the      output  generated  by  the  ANN  by  the  training  data  set  and  the  test  value.       For  supervised  learning  divide  training  data  set  into  train  and  test  segments.     ANN  learns  on  the  training  data  set  and  tests  this  training  on  the  test  dataset.  
  • 8. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Types  of  Training:  Unsupervised   Usually  different  network  structure.  For  example  a  Kohonen  Self   Organizing  Map:   All  inputs  connect  to  ALL  outputs.  Can  have  hidden  layers.  Acts  as  a  classifier,   produces  a  map  (similar  to  segmenta2on)  of  the  input  data.  
  • 9. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Types  of  Training:  Unsupervised   Example  of  Kohonen  Map:  
  • 10. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Types  of  Training:  Unsupervised   Select  an  input  at  random     Find  which  ar2ficial  neuron  is  most  similar   based  on  weight  value.     Adjust  values  of  “winner”  neuron  to  be  close  to     Input.     Keep  doing  this.     Then  you  can  map  a  new  input  “vector”  and     get  classifica2on.  
  • 11. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   •  Lots  of  arithme2c.   –  Video  games  GPU’s  have  helped  alleviate  this  problem.   •  Complex   –  As  bigger  memory  capacity  became  available  bigger   networks  feasible.   •  Coding  of  input.   •  Architecture  selec2on.   –  Thousands  of  architectures  available.     ANN:  Challenges  
  • 12. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   ANN:  New  and  Interes+ng   •  “Genera2ve”  Networks.   –  Produce  new  output  rather  than  classifica2on  and   predic2on.   •  Large  scale  networks  for  text  analysis.   –  Google  word2vec  and  doc2vec.  
  • 13. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   ANN:  LSTM  Recurrent  ANN     •  Long  Term  Short  Term  Memory  ANN.   –  Used  to  generate  a  short  movie.   –  Trained  on  scripts  of  Scifi  movies/TV  shows  on  interweb.   –  (hDp://arstechnica.com/the-­‐mul2verse/2016/06/an-­‐ai-­‐ wrote-­‐this-­‐movie-­‐and-­‐its-­‐strangely-­‐moving/)   –  AI  came  up  with  its  own  name,  Benjamin,  in  an  interview.   –  Biologically  Plausible  ANN,  uses  cogni2ve  psychology   constructs.  
  • 14. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   ANN:  Doc2vec   •  Word2vec/doc2vec   –  Take  a  word  or  document   –  Create  a  “vector”  (list  of  numbers)  300+  long.   –  All  vectors  are  fixed  length   –  Vectors  can  be  compared   •  word2vec  can  show  “odd  man  out”   –  “cat,  dog,  dolphin,  chair”   –  Uses  “distance”  between  vectors  to  decide  odd  man  out  
  • 15. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   ANN:  Doc2vec   •  Doc2vec  does  the  same  thing  for  documents.   –  Converts  documents  to  fixed  length  vectors.   –  Can  calculate  the  similarity  of  documents.   –  Can  be  interrogated  as  in  “which  color  is  best”   •  Prints  out  most  similar  documents  to  input  .   •  Input  phrase  does  not  have  to  be  in  text  corpus  trained  with.   •  Encodes  seman2cs.   •  Poten2al  for  mapping  and  exploring  text   •  Computa2onally  intensive  !  
  • 16. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Thank  You!         www.masscogni2on.com     apj@masscogni2on.com  
  • 17. An  Introduc+on  to  Neural  Networks  –  Agility,  Automa+on  &  AI   Andrew  Jeavons,  Mass  Cogni2on,  2016   Q  &  A   Ray  Poynter   The  Future  Place   Andrew  Jeavons   Mass  Cogni2on