Save 37% off PRO during our Black Friday Sale! »

Serverless Machine Learning Workshop [Tokyo]

Serverless Machine Learning Workshop [Tokyo]

What is Machine Learning and how does it work? But even more importantly, what problems can ML solve for you and your company?

Once you have understood the potential use cases, we will briefly describe the main challenges in the world of Big Data.

Why is deploying ML models so hard and how can Cloud Computing help?

Many MLaaS options are available on the market (AWS, Google, Azure, BigML, etc.). We will see how they compare to each other and which may best fit your needs.

Whenever MLaaS is not enough, you can build your own ML models. We will briefly explain why Serverless is a great deployment strategy for this use case and what problems and limitation arise with it.

Furthermore, we will put these ideas into practice and build a model for Sentiment Analysis, based on Python (scikit-learn), and trained with a public dataset by Stanford University.

36ca8ae7e19067021d0e39c0b72acc2a?s=128

Alex Casalboni

October 04, 2016
Tweet

Transcript

  1. [特別企画]   Serverless  Machine  Learning   Workshop clda.co/serverless-­‐workshop 10/04/2016 東京

  2. @alex_casalboni clda.co/serverless-­‐workshop Workshop  |  東京 Web  Developer  (6+  years) Sr.

     SoBware  Engineer  @  Cloud  Academy Master  in  Computer  Science About  me
  3. What  is  Machine  Learning? Back  to  1959  (Arthur  Samuel) How

     computers  learn  from  Data clda.co/serverless-­‐workshop Workshop  |  東京 How  to  solve  decision  problems
  4. Machine  Learning  pipeline Training Predic6on batch real-­‐Ame Feature   extrac6on

    batch data informaPon features ML  models clda.co/serverless-­‐workshop Workshop  |  東京
  5. ? Machine  Learning  taxonomy Supervised     Learning Unsupervised  

      Learning clda.co/serverless-­‐workshop Workshop  |  東京
  6. ? Machine  Learning  taxonomy classifica8on regression 170
 cm Supervised  

      Learning Unsupervised     Learning clda.co/serverless-­‐workshop Workshop  |  東京
  7. Machine  Learning  taxonomy Supervised     Learning Unsupervised    

    Learning clda.co/serverless-­‐workshop Workshop  |  東京
  8. Machine  Learning  taxonomy clustering rule  extrac8on group A group B

    A, B C Supervised     Learning Unsupervised     Learning clda.co/serverless-­‐workshop Workshop  |  東京
  9. What  problems  can  ML  solve  for  you? Supervised    

    Learning Unsupervised     Learning classifica'on regression clustering rule  extrac'on ? 170
 cm gro gro A, B C clda.co/serverless-­‐workshop Workshop  |  東京
  10. What  problems  can  ML  solve  for  you? Supervised    

    Learning Unsupervised     Learning classifica'on regression clustering rule  extrac'on ? fraud  detecPon 170
 cm gro gro A, B C price  of  a  stock  over  Pme purchase  likelihood user  segmentaPon clda.co/serverless-­‐workshop Workshop  |  東京
  11. Learning Data Machine Cloud Big Science Information Internet Statistics Technology

    Python Future Mining Social Deep IOT Algorithms Management Storage Petabytes Parallel Network Privacy Million NoSQL PaaS SQL Database Exabytes Billion Dataset Hadoop R clda.co/serverless-­‐workshop Workshop  |  東京
  12. Generated  data  started  growing  ~10  years  ago… “90%  of  the

     data  in  the  world  today  has  been     created  in  the  last  two  years  alone”  -­‐  IBM “300+  hours  worth  of  video  content  is  being     uploaded  to  the  site  every  minute”  -­‐  Youtube clda.co/serverless-­‐workshop Workshop  |  東京
  13. …  and  it  keeps  geKng  bigger! clda.co/serverless-­‐workshop Workshop  |  東京

  14. Big  data  challenges Manual  exploraPon  is  not  an  opPon Data-­‐driven

     decisions  are  a  must Distributed/parallel  compuPng The  curse  of  dimensionality clda.co/serverless-­‐workshop Workshop  |  東京
  15. clda.co/serverless-­‐workshop Workshop  |  東京

  16. clda.co/serverless-­‐workshop Workshop  |  東京

  17. Why  is  deploying  ML  models  a  challenge? clda.co/serverless-­‐workshop Workshop  |

     東京
  18. Why  is  deploying  ML  models  a  challenge? 1.  Prototyping  !=

     ProducPon-­‐ready 2.  We  need  ElasPcity 4.  MulP-­‐model  architectures 3.  Too  many  nice-­‐to-­‐have  features clda.co/serverless-­‐workshop Workshop  |  東京 5.  Avoid  lack  of  ownership
  19. Machine  Learning  as  a  Service  (MLaaS) Amazon
 Machine  Learning Azure


    Machine  Learning Google
 PredicAon  API IMB
 Watson  AnalyAcs BigML
 Workshop  |  東京 clda.co/serverless-­‐workshop
  20. Amazon  Machine  Learning AmazonML One  of  the  first  MLaaS  soluPons

     (1  year  old) Great  service  for  classificaPon  and  regression Only  linear  models  (linear  &  logisPc  regression  +  SGD) No  support  for  advanced  scenarios  yet   Workshop  |  東京 clda.co/serverless-­‐workshop
  21. AmazonML  @  Cloud  Academy Workshop  |  東京 clda.co/serverless-­‐workshop clda.co/7-­‐day-­‐free (no

     credit  card  required!)
  22. Serverless  compuAng  to  the  rescue! Transparent  scalability,  elasPcity  and  availability

    Developer-­‐friendly  maintenance  (versioning  +  aliases) AWS   Lambda Event-­‐driven  approach  &  never  pay  for  idle 1  funcPon  =  1  model A/B  tesPng  via  composiPon clda.co/serverless-­‐workshop Workshop  |  東京
  23. clda.co/serverless-­‐workshop Workshop  |  東京

  24. Quick  Example clda.co/ML-­‐Lambda clda.co/serverless-­‐workshop Workshop  |  東京

  25. Thank  you  for  aZending  :) cloudacademy.com Q  &  A 10/04/2016

    東京