Tweet
Tweet

Slide 1

Slide 1 text

Here Come the Robots Django & Machine Learning Tom Dyson, DjangoCon US, October 2018

Slide 2

Slide 2 text

Machine Learning != Artifical Intelligence

Slide 3

Slide 3 text

No content

Slide 4

Slide 4 text

Life 1.0 ➔ Hardware evolves ➔ Software evolves

Slide 5

Slide 5 text

Life 2.0 ➔ Designs its own software ➔ Hardware evolves

Slide 6

Slide 6 text

Life 2.1 ➔ Designs its own software ➔ Hardware evolves (with minor upgrades)

Slide 7

Slide 7 text

Life 3.0 ➔ Designs its own software ➔ Designs its own hardware

Slide 8

Slide 8 text

On the cusp ➔ Life 1.0 – 4 billion years ago ➔ Life 2.0 – 100 thousand years ago ➔ Life 3.0 – 30 years away

Slide 9

Slide 9 text

Machine learning Classical programming uses rules and data to produce answers. Machine Learning uses data and answers to produce rules. François Chollet

Slide 10

Slide 10 text

No content

Slide 11

Slide 11 text

Image recognition What is this a picture of?

Slide 12

Slide 12 text

No content

Slide 13

Slide 13 text

Demo

Slide 14

Slide 14 text

Practical applications ➔ ➔ Content management ➔ Accessibility ➔ NSFW ➔ Language learning

Slide 15

Slide 15 text

Sentiment analysis What is the author feeling?

Slide 16

Slide 16 text

No content

Slide 17

Slide 17 text

Demo

Slide 18

Slide 18 text

Practical applications ➔ Customer support ➔ Survey tools ➔ Fake reviews ➔ News analysis ➔ Better bots

Slide 19

Slide 19 text

Entity extraction What proper nouns is this text about?

Slide 20

Slide 20 text

No content

Slide 21

Slide 21 text

Demo

Slide 22

Slide 22 text

CCA video

Slide 23

Slide 23 text

Practical applications ➔ Content management ➔ Plagiarism tests ➔ Trend spotting

Slide 24

Slide 24 text

Outcome prediction What will happen next, given what we know about the past?

Slide 25

Slide 25 text

Building a model ➔ Prepare ➔ Train ➔ Evaluate ➔ Use

Slide 26

Slide 26 text

No content

Slide 27

Slide 27 text

No content

Slide 28

Slide 28 text

No content

Slide 29

Slide 29 text

No content

Slide 30

Slide 30 text

Demo

Slide 31

Slide 31 text

Practical applications ➔ Audience segmentation ➔ Donation refinement ➔ “You may also like” ➔ Happiness optimisation

Slide 32

Slide 32 text

No content

Slide 33

Slide 33 text

No content

Slide 34

Slide 34 text

“They literally wanted it to be an engine where I’m going to give you 100 résumés, it will spit out the top five, and we’ll hire those.” But by 2015, the company realized its new system was not rating candidates for software developer jobs and other technical posts in a gender-neutral way.

Slide 35

Slide 35 text

That is because Amazon’s computer models were trained to vet applicants by observing patterns in résumés submitted to the company over a 10-year period. Most came from men, a reflection of male dominance across the tech industry. In effect, Amazon’s system taught itself that male candidates were preferable. It penalized résumés that included the word “women’s”, as in “women’s chess club captain”.

Slide 36

Slide 36 text

No content

Slide 37

Slide 37 text

Next steps ➔ If you want to learn ML ◆ Deep Learning with Python ◆ Kaggle ➔ If you want to do something with ML ◆ Read the cloud service docs ➔ Build something amazing