Download as ODP, PPTX
Uploaded byTom Maiaroto
MongoDB & Machine Learning
The document provides an overview of machine learning, detailing its algorithms such as decision trees, random forests, and neural networks. It discusses the applications of machine learning in areas like spam filtering, auto-tagging, and sentiment analysis, highlighting the various programming languages used, including Java and Python. Additionally, it mentions practical implementations, like the Social Harvest tool, which analyzes social data for insights.
MongoDB & Machine Learning
- 1. Machine Learning TomMaiaroto @shift8creative
- 2. What is MachineLearning?
- 3. Algorithms & ApproachesDecision trees Random forests Artificial neural networks k-NN (nearest neighbour) Naive Bayesian classifier
- 4. Algorithms & ApproachesDecision trees Random forests Artificial neural networks k-NN (nearest neighbour) Naive Bayesian classifier
- 5. So could machinesone day rule the earth?
- 6. So could machinesone day rule the earth? Maybe (ok probably not)
- 7. What can MachineLearning do for Apps? Spam filtering
- 8. What can MachineLearning do for Apps? Auto-tagging
- 9. What can MachineLearning do for Apps? All Sorts of Categorization
- 10. What can MachineLearning do for Apps? Sentiment Analysis
- 11. Languages Commonly UsedJava Java-ML, WEKA, Apache Mahout, many more... Python NLTK, scikit-learn, PyML, a good deal more... C++ libDAI, Armadillo, Orange, tons more... and then some others...
- 12. Languages Commonly Used http://www.mloss.org
- 13.
- 14.
- 15.
- 16.
- 17. Geo-spatial Indexing Didsomeone say nearest neighbour?
- 18. Geo-spatial Indexing Didsomeone say nearest neighbour? Design geeks, imagine the visualizations...
- 19. Replication Store massiveamounts of data
- 20.
- 21. Dedicated databases forcalculations All the obvious benefits.
- 22. Map/Reduce It's thebrain.
- 23. Map/Reduce It's thebrain. It's not just for aggregation.
- 24. Map/Reduce It's thebrain. It's not just for aggregation. It's faster than you might think.
- 25. Map/Reduce It's thebrain. It's not just for aggregation. It's faster than you might think. It runs in the database.
- 26. Map/Reduce In the computer. ..
- 27. Example Time! It'ssimple...Just take this...
- 28. Example Time! It'ssimple...Just take this...
- 29. Example Time! Justkidding... Let's Break Down a Naive Bayes Classifier
- 30. Classification /Naive BayesTraining the System
- 31. Classification /Naive BayesTraining the System Simple... $inc
- 32. Classification /Naive BayesJust Keep Count of Words per Category
- 33. Classification /Naive BayesReduce:
- 34. Classification /Naive BayesReduce:
- 35. Classification /Naive BayesFinalize:
- 36. Classification /Naive BayesFinalize:
- 37. Classification /Naive BayesCall the Command:
- 38. Classification /Naive BayesResults: Can see total words. Can also see word counts per category.
- 39. Classification /Naive BayesResults: ...and of course the scores per category... cae = arts and entertainment cs = science ...
- 40. Classification /Naive BayesAccurate even with little training
- 41. MongoDB on asmall VM Took 1.7 seconds
- 42. Compared to say PHP 33 seconds and timed out
- 43. More training data== exponentially faster than PHP
- 44. Classification /Naive BayesThis wasn't even a full map/reduce
- 45. Your mileage willvary based on formula
- 46. You can cachecertain values for speed
- 47. Don't forget aboutstored JavaScript (but use it wisely)
- 48. Porter Stemming Algorithm Thank You Martin Porter http://tartarus.org/martin/PorterStemmer
- 49. Porter Stemming Algorithm Exists for nearly every language
- 50. MongoDB will useJavaScript of course
- 51.
- 52. Porter Stemming Algorithm About 2.5x faster than PHP class
- 53. 663x faster thana web browser
- 54. Porter Stemming Algorithm About 2.5x faster than PHP class
- 55. 663x faster thana web browser
- 56. 7x slower thanPHP PECL extension
- 57. Real World ApplicationSocial Harvest Analyzes social data from the internet to determine languages spoken, gender, age, sentiment analysis, and categories. www.social-harvest.com
- 58. Real World ApplicationSocial Harvest Who doesn't like pie charts?
- 59.
- 60. Follow Tom @shift8creativewww.shift8creative.com www.social-harvest.com www.union-of-rad.com Thank You!