Machine Learning: Make Your Ruby Code Smarter
1 like209 views
The document discusses concepts in machine learning (ML) and artificial intelligence (AI), explaining the differences between supervised and unsupervised learning, and detailing linear regression and logistic regression. It also covers neural networks, deep learning, and specific applications like recommendation systems and clustering. Additionally, it provides code examples, feature scaling techniques, and resources for further learning.
![require './neural_network'
LOCATIONS = [:home, :work, :tennis, :parents]
LOCATIONS_INDEXED = LOCATIONS.map.with_index { |x, i| [x, i] }.to_h
XX = [
# week 1
# 1st day of week, 8am
[:work, 1, 8], [:tennis, 1, 17], [:home, 1, 20],
[:work, 2, 8], [:home, 2, 18],
[:work, 3, 8], [:tennis, 3, 17], [:home, 3, 20],
[:work, 4, 8], [:home, 4, 18],
[:work, 5, 8], [:home, 5, 18],
[:parents, 7, 13], [:home, 7, 18],
# week 2
[:work, 1, 8], [:home, 1, 18],
[:work, 2, 8], [:home, 2, 18],
[:work, 3, 8], [:tennis, 3, 17], [:home, 3, 20],
[:work, 4, 8], [:home, 4, 18],
[:work, 5, 8], [:home, 5, 18],](https://image.slidesharecdn.com/machine-learning-161208065538/85/Machine-Learning-Make-Your-Ruby-Code-Smarter-75-320.jpg)
![XX.each do |destination, day, time|
yy << LOCATIONS_INDEXED[destination]
xx << [day.to_f/7, time.to_f/24]
end
features scaling](https://image.slidesharecdn.com/machine-learning-161208065538/85/Machine-Learning-Make-Your-Ruby-Code-Smarter-76-320.jpg)
![[
[1, 16.5], [1, 17], [1, 17.5], [1, 17.8],
[2, 17], [2, 18.1],
[4, 18],
[6, 23],
[7, 13],
].each do |day, time|
res = nn.predict_with_probabilities([
[day.to_f/7, time.to_f/24]
]).first.
select {|v| v[0] > 0} # filter zero probabilities
puts "#{day} #{time} t #{res.map {|v| [LOCATIONS[v[1]], v[0]]}.inspect}"
end](https://image.slidesharecdn.com/machine-learning-161208065538/85/Machine-Learning-Make-Your-Ruby-Code-Smarter-79-320.jpg)
![1 16.5 [[:tennis , 0.97]]
1 17 [[:tennis , 0.86], [:home , 0.06]]
1 17.5 [[:home , 0.52], [:tennis, 0.49]]
1 17.8 [[:home , 0.82], [:tennis, 0.22]]
2 17 [[:tennis , 0.85], [:home , 0.06]]
2 18.1 [[:home , 0.95], [:tennis, 0.07]]
4 18 [[:home , 0.96], [:tennis, 0.08]]
6 23 [[:home , 1.00]]
[:work, 1, 8], [:tennis, 1, 17], [:home, 1, 20],
[:work, 2, 8], [:home, 2, 18],
[:work, 3, 8], [:tennis, 3, 17], [:home, 3, 20],
[:work, 4, 8], [:home, 4, 18],
[:work, 5, 8], [:home, 5, 18],
[:parents, 7, 13], [:home, 7, 18],
# week 2
[:work, 1, 8], [:home, 1, 18],
[:work, 2, 8], [:home, 2, 18],
[:work, 3, 8], [:tennis, 3, 17], [:home, 3, 20],
[:work, 4, 8], [:home, 4, 18],
[:work, 5, 8], [:home, 5, 18],](https://image.slidesharecdn.com/machine-learning-161208065538/85/Machine-Learning-Make-Your-Ruby-Code-Smarter-80-320.jpg)
Machine Learning: Make Your Ruby Code Smarter
- 4. awesome web and mobile apps since 2005
- 11. terms
- 12. AI (artificial intelligence) - the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages
- 13. ML (machine learning) - is a type of artificial intelligence (AI) that provides computers with the ability to learn without being explicitly programmed. Machine learning focuses on the development of computer programs that can teach themselves to grow and change when exposed to new data.
- 14. without being explicitly programmed
- 15. FF NN cost function
- 16. FF NN Cost Function I’m kidding
- 17. cost function with regularization
- 18. 2 types of ML supervised learning unsupervised learning
- 19. supervised the training data is labeled, eg. we know the correct answer
- 20. unsupervised the training data is not labeled, eg. we would figure out hidden correlations by ourselves
- 22. (x(1) , y(1) ), (x(2) , y(2) )…(x(m) , y(m) ) m training examples of (x(i) , y(i) ) x(i) - feature y(i) - label x(i) y(i)
- 23. training set learning algorithm hθ(x)x(new data) y(prediction)
- 24. training set learning algorithm hθ(x)x(new data) y(prediction)
- 26. (x, y) y = hθ(x) = θ0 + θ1x find θ0 and θ1
- 27. hθ(x) = θ0 + θ1x1 + θ2x2 + …+ θnxn many features, n - number of features
- 28. size, sq.m x1 # rooms x2 age x3 price y 80 3 22 2.9M 90 4 24 3.1M 75 3 28 2.5M 110 5 20 3.3M
- 29. 1 USD = 3.85 NIS
- 30. hθ(x) = θ0 + θ1x1 summate the prediction error on training set
- 31. Linear Regression Cost Function
- 32. minimize J(θ) funding a minimum of cost function = “learning”
- 33. gradient descent batch, stochastic, etc, or advanced optimization algorithms to find a global (sometimes local) minimum of cost function J 𝞪 - learning rate, a parameter of gradient descent
- 34. (x(1) , y(1) ), (x(2) , y(2) )…(x(m) , y(m) ) gradient descent θ0, θ1, θ2, …, θn magic inside
- 35. hθ(x) = θ0 + θ1x1 + θ2x2 + …+ θnxn we’re ready to predict
- 36. features scaling 0 ≤ x ≤ 1
- 37. size, sq.m size, sq.m / 110 x1 80 0.72 90 0.81 75 0.68 110 1
- 38. mean normalization average value of the feature is ~0 -0.5 ≤ x ≤ 0.5
- 39. size, sq.m (size, sq.m / 110) - 0.8025 x1 80 -0.0825 90 0.075 75 -0.1226 110 0.1975
- 40. matrix manipulations X = n x 1 vector, ϴ = n x 1 vector hθ(x) = θ0 + θ1x1 + θ2x2 + …+ θnxn hθ(x) = ϴT X
- 41. GPU
- 44. classifier
- 45. y = 1, true y = 0, false
- 46. hθ(x) = g(ϴT X) hθ(X) - estimated probability that y = 1 on input X g(z) - logistic non-linear function
- 47. logistic function g(z) there is a few: sigmoid, tahn, ReLUs, etc image source: Wikipedia
- 48. (x(1) , y(1) ), (x(2) , y(2) )…(x(m) , y(m) ) minimize the cost function vector θ y = {0, 1}
- 49. training set learning algorithm hθ(x)x(new data) y(prediction) hθ(x) = g(ϴT X) y ≥ 0.5 - true y < 0.5 - false
- 52. y = 1, true y = 0, false y = 0, false
- 53. don’t implement it at home use libsvm, liblinear, and others
- 56. feed forward neural network input layer hidden layer output layer
- 59. logistic unit x0 x1 x2 θ1 θ2 θ3 hθ = g(x0θ0 + x1θ1 + x2θ2) θ - weights g - activation function
- 60. logistic function g(z) there is a few: sigmoid, tahn, ReLUs, etc image source: Wikipedia
- 61. output: probabilities 0.4765 that y = 2 0.7123 that y = 1
- 62. net with no hidden layers no hidden layers = one-vs-all logistic regression
- 63. cost function sometimes called loss function of NN, a representation of an error between a real and a predicted value
- 64. training set learning algorithm θx(new data) y(prediction)
- 65. backprop backward propagation of errors
- 66. gradient descent + backprop “deep learning” - is training a neural net “deep” - because we have many layers
- 67. convolutional neural nets widely used for image processing and object recognition
- 68. recurrent neural nets widely used for natural language processing
- 72. 2016
- 74. tangledpath/ruby-fann Ruby library for interfacing with FANN (Fast Artificial Neural Network)
- 75. require './neural_network' LOCATIONS = [:home, :work, :tennis, :parents] LOCATIONS_INDEXED = LOCATIONS.map.with_index { |x, i| [x, i] }.to_h XX = [ # week 1 # 1st day of week, 8am [:work, 1, 8], [:tennis, 1, 17], [:home, 1, 20], [:work, 2, 8], [:home, 2, 18], [:work, 3, 8], [:tennis, 3, 17], [:home, 3, 20], [:work, 4, 8], [:home, 4, 18], [:work, 5, 8], [:home, 5, 18], [:parents, 7, 13], [:home, 7, 18], # week 2 [:work, 1, 8], [:home, 1, 18], [:work, 2, 8], [:home, 2, 18], [:work, 3, 8], [:tennis, 3, 17], [:home, 3, 20], [:work, 4, 8], [:home, 4, 18], [:work, 5, 8], [:home, 5, 18],
- 76. XX.each do |destination, day, time| yy << LOCATIONS_INDEXED[destination] xx << [day.to_f/7, time.to_f/24] end features scaling
- 77. 2 ➞ 25 ➞ 4 one hidden layer with 25 units
- 78. 100% accuracy on training set
- 79. [ [1, 16.5], [1, 17], [1, 17.5], [1, 17.8], [2, 17], [2, 18.1], [4, 18], [6, 23], [7, 13], ].each do |day, time| res = nn.predict_with_probabilities([ [day.to_f/7, time.to_f/24] ]).first. select {|v| v[0] > 0} # filter zero probabilities puts "#{day} #{time} t #{res.map {|v| [LOCATIONS[v[1]], v[0]]}.inspect}" end
- 80. 1 16.5 [[:tennis , 0.97]] 1 17 [[:tennis , 0.86], [:home , 0.06]] 1 17.5 [[:home , 0.52], [:tennis, 0.49]] 1 17.8 [[:home , 0.82], [:tennis, 0.22]] 2 17 [[:tennis , 0.85], [:home , 0.06]] 2 18.1 [[:home , 0.95], [:tennis, 0.07]] 4 18 [[:home , 0.96], [:tennis, 0.08]] 6 23 [[:home , 1.00]] [:work, 1, 8], [:tennis, 1, 17], [:home, 1, 20], [:work, 2, 8], [:home, 2, 18], [:work, 3, 8], [:tennis, 3, 17], [:home, 3, 20], [:work, 4, 8], [:home, 4, 18], [:work, 5, 8], [:home, 5, 18], [:parents, 7, 13], [:home, 7, 18], # week 2 [:work, 1, 8], [:home, 1, 18], [:work, 2, 8], [:home, 2, 18], [:work, 3, 8], [:tennis, 3, 17], [:home, 3, 20], [:work, 4, 8], [:home, 4, 18], [:work, 5, 8], [:home, 5, 18],
- 81. borisnadion/suggested-destination-demo ruby code of the demo
- 82. tensorflow but you will need to learn Python
- 89. Jane Arthur John Star Wars VII 5 5 1 Dr. Strange 5 5 ? Arrival 5 ? 1
- 90. automatic features and their weights detection based on the user votes
- 91. similarity between users and between items
- 92. what to google