Jump to: navigation, search

Difference between revisions of "Meteos"

(→‎Projects)
(→‎Meteos (Machine Learning as a Service))
 
(29 intermediate revisions by one other user not shown)
Line 1: Line 1:
−
== Meteos (Machine Learning as a Service) ==
+
== Meteos (Machine Learning as a Service) (Deprecated) ==
 +
(check the new design and mission of [[Meteos/Meteos2.0]])
 +
 
 
Meteos is Machine Learning as a Service (MLaaS) in Apache Spark.
 
Meteos is Machine Learning as a Service (MLaaS) in Apache Spark.
  
 
Meteos allows users to analyze huge amount of data and predict a value by data mining and machine learning algorithms.
 
Meteos allows users to analyze huge amount of data and predict a value by data mining and machine learning algorithms.
 
Meteos create a workspace of Machine Learning via sahara spark plugin and manage some resources and jobs regarding Machine Learning.
 
Meteos create a workspace of Machine Learning via sahara spark plugin and manage some resources and jobs regarding Machine Learning.
 +
 +
Meteos is named from Meteo (Meteorologist) + OS (OpenStack).
 +
 +
[[File:Meteos-overview.png]]
  
 
== Projects ==
 
== Projects ==
  
−
=== Meteos ===
+
==== Meteos ====
  
 
{| border="1" cellpadding="2"
 
{| border="1" cellpadding="2"
Line 20: Line 26:
 
|}
 
|}
  
−
=== Python Meteos Client ===
+
==== Python Meteos Client ====
 
{| border="1" cellpadding="2"
 
{| border="1" cellpadding="2"
 
| Source code
 
| Source code
Line 32: Line 38:
 
|}
 
|}
  
−
=== Getting Started with Meteos ===
+
==== Meteos UI ====
  
−
Instructions for getting started with Meteos using Devstack are available at: [[Meteos/Devstack|Meteos on Devstack]]
+
{| border="1" cellpadding="2"
 +
| Source code
 +
| https://github.com/openstack/meteos-ui
 +
|-
 +
| Bug tracker
 +
| https://bugs.launchpad.net/meteos-ui
 +
|-
 +
| Feature tracker
 +
| https://blueprints.launchpad.net/meteos-ui
 +
|}
  
−
== Use Cases ==
+
==== IRC ====
 +
http://webchat.freenode.net/?channels=openstack-meteos
  
−
Machine Learning consists of the following phases.
+
== Design & Use Cases ==
  
−
* Learning Phase - Analyze huge amounts of data and create a Prediction Model
+
* [[Meteos/Architecture| Meteos Architecture]]
−
* Prediction Phase - Predict a value according to the input value by using Prediction Model
 
  
−
=== Use Cases in Learning Phase ===
+
* [[Meteos/Usecases| Meteos Use Cases]]
  
−
* Upload Raw Data - Upload a raw data to Object Stroage
+
* [[Meteos/DatasetsandModels| Meteos Dataset and Prediction Model]]
−
* Parse Raw Data - Parse a raw data to enable MLlib (Apache Spark's scalable machine learning library) to handle it. Users are allowed to parse a parsed data again.
 
−
* Create Prediction Model - Create a Prediction Model by using MLlib.
 
  
−
=== Use Cases in Prediction Phase ===
+
== Getting Started with Meteos ==
  
−
* Predict - Input any value and retrieve predicted value.
+
Instructions for getting started with Meteos using Devstack are available at: [[Meteos/Devstack|Meteos on Devstack]]
−
 
 
−
==== [[Meteos/Usecase|Use Case (Predict sales using Meteos)]] ====
 
−
 
 
−
== Archtecture (TBD) ==
 
−
 
 
−
Meteos consist of meteos-api service and meteos-engine service.
 
−
 
 
−
* meteos-api - web service which has REST interface.
 
−
 
 
−
* meteos-engine - service which manage Meteos resources.
 
−
 
 
−
[[File:Meteos-architecture.png]]
 
−
 
 
−
== Resource (TBD) ==
 
−
 
 
−
* Experiment Template - Template which define experiment (e.g. number of master/worker nodes, spark version, base VM image, flavor, network, ...)
 
−
 
 
−
* Experiment - a workspace of Machine Learning
 
−
 
 
−
* Data Set - a data parsed by user to create a Prediction Model
 
−
 
 
−
* Prediction Model - a model produced by data mining and machine learning algorithms
 
−
 
 
−
* Learning Job - a job which consists of input data, output data(predicted data), job status, job stdout/stderr.
 
−
 
 
−
== API (TBD) ==
 
−
 
 
−
=== Experiment Template ===
 
−
 
 
−
* Create Experiment Template
 
−
** POST /v1/<tenant_id>/templates
 
−
 
 
−
* List Experiment Templates
 
−
** GET /v1/<tenant_id>/templates
 
−
 
 
−
* Show Experiment Template
 
−
** GET /v1/<tenant_id>/templates/<template_id>
 
−
 
 
−
* Update Experiment Template
 
−
** PUT /v1/<tenant_id>/templates
 
−
 
−
* Delete Experiment Template
 
−
** DELETE /v1/<tenant_id>/templates/<template_id>
 
−
 
 
−
=== Experiment ===
 
−
 
 
−
* Create Experiment
 
−
** POST /v1/<tenant_id>/experiments
 
−
 
 
−
* List Experiments
 
−
** GET /v1/<tenant_id>/experiments
 
−
 
 
−
* Show Experiment
 
−
** GET /v1/<tenant_id>/experiments/<experiment_id>
 
−
 
 
−
* Update Experiment
 
−
** PUT /v1/<tenant_id>/experiments
 
−
 
−
* Delete Experiment
 
−
** DELETE /v1/<tenant_id>/experiments/<experiment_id>
 
−
 
 
−
=== Data Set ===
 
−
 
 
−
* Create Data Set
 
−
** POST /v1/<tenant_id>/datasets
 
−
 
 
−
* List Data Sets
 
−
** GET /v1/<tenant_id>/datasets
 
−
 
 
−
* Show Data Sets
 
−
** GET /v1/<tenant_id>/datasets/<dataset_id>
 
−
 
 
−
* Update Data Set
 
−
** PUT /v1/<tenant_id>/datasets
 
−
 
−
* Delete Data Set
 
−
** DELETE /v1/<tenant_id>/datasets/<dataset_id>
 
  
−
==== Data Set Actions====
+
== API ==
  
−
* Export Data Set to Object Storage
+
* [[Meteos/API| Meteos API]]
−
** POST /v1/<tenant_id>/datasets/<dataset_id>/action
 
−
*** BODY {"export"}
 
  
−
=== Prediction Model ===
+
== Examples (CLI) ==
  
−
* Create Prediction Model
+
* [[Meteos/ExampleLinear| Predict a Sales Figures by using LinearRegression Model]]
−
** POST /v1/<tenant_id>/models
 
  
−
* List Prediction Models
+
* [[Meteos/ExampleDecisionTree| Make a Decision to buy a stock by using DecisionTree Model]]
−
** GET /v1/<tenant_id>/models
 
  
−
* Show Model
+
* [[Meteos/ExampleKmeans| Classify Users by Skill using Kmeans Model]]
−
** GET /v1/<tenant_id>/models/<moded_id>
 
  
−
* Update Model
+
* [[Meteos/ExampleRecommend| Recommend a Movie by using Recommendation Model]]
−
** PUT /v1/<tenant_id>/models
 
  
−
* Delete Model
+
* [[Meteos/ExampleWord2Vec| Search Synonyms by using Word2Vec Model]]
−
** DELETE /v1/<tenant_id>/models/<model_id>
 
  
−
==== Prediction Model Actions====
+
== Examples (GUI) ==
  
−
* Export Prediction Model to Object Storage
+
* [[Meteos/ExampleNativebyes| Detect a Spam Mail using Native byes Model]]
−
** POST /v1/<tenant_id>/models/<model_id>/action 
 
−
*** BODY {"export"}
 
  
−
=== Learning Job ===
+
== Tips ==
  
−
* Create Learning Job
+
* [[Meteos/Howto| How to increase the model accuracy]]
−
** POST /v1/<tenant_id>/jobs
 
  
−
* List Learning Jobs
+
== Boston Summit ==
−
** GET /v1/<tenant_id>/jobs
 
  
−
* Show Learning Job
+
* PPT Slide
−
** GET /v1/<tenant_id>/jobs/<job_id>
+
https://www.slideshare.net/guchi_hiro/openstack-meteos-machine-learning-as-a-service
  
−
* Delete Learning Job
+
* Demo
−
** DELETE/v1/<tenant_id>/jobs/<job_id>
+
https://www.youtube.com/watch?v=YFv6SQvqNFs

Latest revision as of 04:04, 8 July 2018

Meteos (Machine Learning as a Service) (Deprecated)

(check the new design and mission of Meteos/Meteos2.0)

Meteos is Machine Learning as a Service (MLaaS) in Apache Spark.

Meteos allows users to analyze huge amount of data and predict a value by data mining and machine learning algorithms. Meteos create a workspace of Machine Learning via sahara spark plugin and manage some resources and jobs regarding Machine Learning.

Meteos is named from Meteo (Meteorologist) + OS (OpenStack).

Meteos-overview.png

Projects

Meteos

Source code https://github.com/openstack/meteos
Bug tracker https://bugs.launchpad.net/meteos
Feature tracker https://blueprints.launchpad.net/meteos

Python Meteos Client

Source code https://github.com/openstack/python-meteosclient
Bug tracker https://bugs.launchpad.net/python-meteosclient
Feature tracker https://blueprints.launchpad.net/python-meteosclient

Meteos UI

Source code https://github.com/openstack/meteos-ui
Bug tracker https://bugs.launchpad.net/meteos-ui
Feature tracker https://blueprints.launchpad.net/meteos-ui

IRC

http://webchat.freenode.net/?channels=openstack-meteos

Design & Use Cases

Getting Started with Meteos

Instructions for getting started with Meteos using Devstack are available at: Meteos on Devstack

API

Examples (CLI)

Examples (GUI)

Tips

Boston Summit

  • PPT Slide

https://www.slideshare.net/guchi_hiro/openstack-meteos-machine-learning-as-a-service

  • Demo

https://www.youtube.com/watch?v=YFv6SQvqNFs

Retrieved from "https://wiki.openstack.org/w/index.php?title=Meteos&oldid=162583"