Overview

Machine Learning

Artificial Neural Networks - ANN’s

Deep Learning

Computer Vision - CV

Natural Language Processing - NLP

Why Big Data?

Machine Learning

Explicitly programmed:

power <- function(a,b) {
  c <- a^b # known function
  return(c)
}
power(1,3)
## 1
power(2,3)
## 8
power(3,4)
## 81

Machine Learning

What if we don’t know what function to apply, but we have a lot of examples?

func(2)
## 2.895133
func(1)
## 1.666248
func(6)
## 6.399109

Machine Learning

Machine learning is an application of artificial intelligence (AI) that provides systems the ability to automatically learn and improve from experience without being explicitly programmed.

  • Supervised machine learning algorithms
  • Unsupervised machine learning algorithms
  • Reinforcement machine learning algorithms

Supervised machine learning

Unsupervised machine learning

Reinforcement machine learning

Artificial Neural Network - ANN

Single-layer Perceptron

  • Input Nodes (input layer)
  • Connections and Weights
  • Activation Function
  • Output Nodes (output layer)
## 
## Attaching package: 'matrixStats'
## The following object is masked from 'package:plyr':
## 
##     count

Artificial Neural Network - ANN

Single-layer Perceptron

Let’s try and predict rating:

calories protein fat sodium fiber rating
70 4 1 130 10.0 68
120 3 5 15 2.0 34
70 4 1 260 9.0 59
50 4 0 140 14.0 94
110 2 2 180 1.5 30

Artificial Neural Network - ANN

Single-layer Perceptron

Artificial Neural Network - ANN

Single-layer Perceptron

calories protein fat sodium fiber rating pred_rating
120 3 5 15 2.0 34 31
70 4 1 260 9.0 59 70
110 2 2 180 1.5 30 34
90 2 1 200 4.0 49 47
120 1 2 220 0.0 18 22

Artificial Neural Network - ANN

Deep Learning