Machine Learning Real Life Applications By Examples - Mario Cartia

1. DATA DRIVEN INNOVATION Rome 2017 | Open Summit MACHINE LEARNING REAL LIFE APPLICATIONS BY EXAMPLES SPEAKER MARIO CARTIA MARIO@BIG-DATA.NINJA

2. DDI R O M E| 2017 M A RI O C A RTI A Can machines think? Computing machinery and intelligence. Mind, 59, 433-460 (1950) Turing A.M.

3. DDI R O M E| 2017 M A RI O C A RTI A 1968 2001: A Space Odyssey “I'm sorry Dave, I'm afraid I can't do that”

4. DDI R O M E| 2017 M A RI O C A RTI A 1982 Supercar

5. DDI R O M E| 2017 M A RI O C A RTI A 1983 Wargames

6. DDI R O M E| 2017 M A RI O C A RTI A 1996 Kasparov vs. Deep Blue

7. DDI R O M E| 2017 M A RI O C A RTI A Does Deep Blue use artificial intelligence? The short answer is "no." Earlier computer designs that tried to mimic human thinking weren't very good at it. No formula exists for intuition. So Deep Blue's designers have gone "back to the future." Deep Blue relies more on computational power and a simpler search and evaluation function. The long answer is "no." "Artificial Intelligence" is more successful in science fiction than it is here on earth, and you don't have to be Isaac Asimov to know why it's hard to design a machine to mimic a process we don't understand very well to begin with. Source: https://www.research.ibm.com/deepblue/meet/html/d.3.3a.shtml

8. DDI R O M E| 2017 M A RI O C A RTI A Decision Tree (IF... THEN)

9. DDI R O M E| 2017 M A RI O C A RTI A “Machine learning is the subfield of computer science that gives computers the ability to learn without being explicitly programmed.” Arthur Samuel, 1959

10. DDI R O M E| 2017 M A RI O C A RTI A Spam Email Filtering

11. DDI R O M E| 2017 M A RI O C A RTI A Email Category Tabs

12. DDI R O M E| 2017 M A RI O C A RTI A “If you can't explain it simply you don't understand it well enough”

13. DDI R O M E| 2017 M A RI O C A RTI A SUPERVISED LEARNING Supervised learning is where you have input variables (x) and an output variable (y) and you use an algorithm to learn the mapping function from the input to the output y=f(x)

14. DDI R O M E| 2017 M A RI O C A RTI A SUPERVISED LEARNING Classification is a general process related to categorization, the process in which ideas and objects are recognized, differentiated, and understood A classification system is an approach to accomplishing classification

15. DDI R O M E| 2017 M A RI O C A RTI A CLASSIFICATION In Machine Learning, Naive Bayes Classifiers are a family of simple probabilistic classifiers based on applying Bayes' theorem with strong (naive) independence assumptions between the features

16. DDI R O M E| 2017 M A RI O C A RTI A NAIVE BAYES CLASSIFIERS Naive Bayes has been studied extensively since the 1950s and remains a popular (baseline) method for text categorization, the problem of judging documents as belonging to one category or the other (such as spam or legitimate, sports or politics, etc.) with word frequencies as the features

17. DDI R O M E| 2017 M A RI O C A RTI A TEXT CATEGORIZATION SUPERVISED LEARNING CLASSIFICATION NAIVE BAYES CLASSIFIER ?

18. DDI R O M E| 2017 M A RI O C A RTI A

23. DDI R O M E| 2017 M A RI O C A RTI A Recommendation system

28. DDI R O M E| 2017 M A RI O C A RTI A UNSUPERVISED LEARNING Unsupervised learning algorithms are machine learning algorithms that work without a desired output label Essentially, the algorithm attempts to estimate the underlying structure of the population of input data

29. DDI R O M E| 2017 M A RI O C A RTI A UNSUPERVISED LEARNING Collaborative Filtering is a method of making automatic predictions (filtering) about the interests of a user by collecting preferences or taste information from many users (collaborating) In the more general sense, Collaborative Filtering is the process of filtering for information or patterns using techniques involving collaboration among multiple agents, viewpoints, data sources, etc.

30. DDI R O M E| 2017 M A RI O C A RTI A COLLABORATIVE FILTERING Applications of Collaborative Filtering typically involve very large data sets As the numbers of users and items grow, traditional CF algorithms will suffer serious scalability problems Large web companies use clusters of machines to scale recommendations for their millions of users

31. DDI R O M E| 2017 M A RI O C A RTI A RECOMMENDATION SYSTEM UNSUPERVISED LEARNING COLLABORATIVE FILTERING USER BASED / ITEM BASED / OTHER ?

36. DDI R O M E| 2017 M A RI O C A RTI A Targeted Advertising

37. DDI R O M E| 2017 M A RI O C A RTI A Targeted Advertising

38. DDI R O M E| 2017 M A RI O C A RTI A UNSUPERVISED LEARNING Cluster analysis or Clustering is the task of grouping a set of objects in such a way that objects in the same group (called a cluster) are more similar (in some sense or another) to each other than to those in other groups (clusters)

39. DDI R O M E| 2017 M A RI O C A RTI A CLUSTERING K-means clustering is a method of vector quantization, originally from signal processing, that is popular for cluster analysis in data mining K-means clustering aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean

40. DDI R O M E| 2017 M A RI O C A RTI A TARGETED ADVERTISING UNSUPERVISED LEARNING CLUSTERING K-MEANS CLUSTERING ?

43. DDI R O M E| 2017 M A RI O C A RTI A TYPICAL ML WORKFLOW ü Data and problem definition ü Data collection ü Data preprocessing ü Data analysis and modeling with unsupervised and supervised learning ü Process evaluation

44. DDI R O M E| 2017 M A RI O C A RTI A EVALUATION METRICS The root-mean-square deviation (RMSD) or root-mean-square error (RMSE) is a frequently used measure of the differences between values (sample and population values) predicted by a model or an estimator and the values actually observed

45. DDI R O M E| 2017 M A RI O C A RTI A BEYOND ML Deep learning is a branch of machine learning based on a set of algorithms that attempt to model high level abstractions in data Deep learning is part of a broader family of machine learning methods based on learning representations

46. DDI R O M E| 2017 M A RI O C A RTI A BEYOND ML One of the promises of Deep Learning is replacing handcrafted features with efficient algorithms for unsupervised or semi-supervised feature learning and hierarchical feature extraction Some of the representations are inspired by advances in neuroscience

47. DDI R O M E| 2017 M A RI O C A RTI A BEYOND ML Various Deep Learning architectures such as deep neural networks have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state- of-the-art results on various tasks

50. DDI R O M E| 2017 M A RI O C A RTI A ML & BIG DATA “We don’t have better algorithms. We just have more data.” Peter Norvig Google’s Research Director

51. DDI R O M E| 2017 M A RI O C A RTI A ML & BIG DATA Apache Hadoop is an open-source software framework used for distributed storage and processing of big data sets using clusters built from commodity hardware

52. DDI R O M E| 2017 M A RI O C A RTI A ML & BIG DATA Apache Spark is a fast and general-purpose cluster computing system It provides high-level APIs in Scala, Java, Python and R, and an optimized engine that supports general execution graphs It also supports a rich set of higher-level tools including Spark SQL for SQL and structured data processing, MLlib for machine learning, GraphX for graph processing, and Spark Streaming

53. DDI R O M E| 2017 M A RI O C A RTI A ML & BIG DATA

54. DDI R O M E| 2017 M A RI O C A RTI A WHY TO USE SCALA? Spark Survey 2016

55. DDI R O M E| 2017 M A RI O C A RTI A WHY TO USE SCALA? Scala is one of the most exciting languages to be created in the 21st century. It is a multi-paradigm language that fully supports functional, object-oriented, imperative and concurrent programming. It also has a strong type system, and from our point of view, strong type is a convenient form of self-documenting code. Scala works on the JVM and has access to the riches of the Java ecosystem, but it is less verbose than Java. As we employ it for ND4J, its syntax is strikingly similar to Python, a language that many data scientists are comfortable with. Like Python, Scala makes programmers happy, but like Java, it is quite fast. Finally, Apache Spark is written in Scala, and any library that purports to work on distributed run times should at the very least be able to interface with Spark Source: https://deeplearning4j.org/scala

56. DDI R O M E| 2017 M A RI O C A RTI A GRAZIE! MARIO@BIG-DATA.NINJA

Machine Learning Real Life Applications By Examples - Mario Cartia

Vous êtes en train de lire un aperçu.

Consultez-les par la suite

Machine Learning Real Life Applications By Examples - Mario Cartia

Plus De Contenu Connexe

Les utilisateurs ont également aimé (17)

Similaire à Machine Learning Real Life Applications By Examples - Mario Cartia (20)

Plus par Data Driven Innovation (20)

Plus récents (20)