MakerSquare Clustering

Clustering lesson.

Transcript

1. Clustering
2. Elben Shira! Software Engineer! Spredfast (Mass Relevance)
3. Please ask questions!
4. Why data structures and algorithms?
 
 So we can build fast, stable systems.
 
 Well-designed systems require wisely-chosen data structures and algorithms.
5. My day-to-day ! Millions requests/minute.
 10,000 Tweets/second.
 
 We think about data structures and algorithms every day.
 
 We want people who know this stuff.
6. Clustering problem.
7. How many clusters do you see? Where are they? Example: T-shirt sizes Height Weight
8. Weight Height Humans are good! Can we make a computer just as good? Example: T-shirt sizes
9. Use-cases of clustering algorithms ! Search engines! ! Pattern recognition! ! Faces.! ! Tracking cars in videos.! ! Bioinformatics! ! Gene expressions in cancer cells.! ! Social graph search! ! Who are your closest (online) friends?!
13. Your Project ! Develop a clustering algorithm to find clusters of metropolitan areas.
14. Demo
15. Input ! numClusters = 3; ! data = [ [123, 145], [152, 190], [170, 184], … ]; Output ! clusters = [ // cluster 0 [[123, 145], …], ! // cluster 1 [[152, 190], …], ! // cluster 2 [[170, 184], …] ];
16. Some tips ! Observe every step of your own thinking. How did your brain conclude those clusters?! ! Play around with different examples. See how your thinking and proposed algorithm would be affected.
18. Don’t write any code until you have a strategy.! ! Use paper and pencil.! ! Progress is possible only if we train ourselves to think about programs without thinking of them as pieces of executable code.! ! — Dijkstra
19. Have Fun! git clone https://bitbucket.org/ elben/makersquare-clustering.git