+ All Categories
Home > Technology > H2O Core Introduction

H2O Core Introduction

Date post: 23-Jan-2018
Category:
Upload: avkash-chauhan
View: 146 times
Download: 0 times
Share this document with a friend
37
H2O Core Architecture & Algorithms Avkash Chauhan [email protected] @avkashchauhan https://www.linkedin.com/in/avkashchauhan
Transcript
Page 1: H2O Core Introduction

H2O Core Architecture &

Algori thms

Avkash [email protected]

@avkashchauhanhttps://www.linkedin.com/in/avkashchauhan

Page 2: H2O Core Introduction

Please visit: http://www.h2o.ai/customers/# H2O Users List: http://www.h2o.ai/user-list/

Page 3: H2O Core Introduction

H2O Platform(s)

In-Memory, Distributed Machine Learning Algorithms with H2O Flow GUI

H2O AI Open Source Engine Integration with Spark

DEEP WATER

Page 4: H2O Core Introduction

Key features

• Open Source (Apache 2.0)

• All supported ML algorithms are coded by our engineers

• Designed for speed, scalability and for super large data-sets

• Same distribution for open source community & enterprise

• Very active production, every other week release

• Vibrant open source community

o https://community.h2o.ai

• Enterprise Support portal

o https://support.h2o.ai

• We have 70,000 users, 8,000 organizations and growing daily

Page 5: H2O Core Introduction

Usage: Simple Solut ion

o Single Deployable compiled Java code (jar)

o Ready to use point and click FLOW Interface

o Connection from R and Python after specific packages are

installed

o Use Java, Scala natively and any other language through

RESTful API

o Deployable models - Binary & Java (POJO & MOJO)

o One click prediction/scoring engine

Page 6: H2O Core Introduction

Usage: Complex Solut ion

o Multi-node Deployment

o Spark and Hadoop distributed environment

• Sparkling Water (Spark + H2O)

o Data ingested from various inputs

• S3, HDFS, NFS, JDBC, Object store etc.

• Streaming support in Spark (through Sparking Water)

o Distributed machine learning for every algorithm in platform

o Prediction service deployment on several machines

Page 7: H2O Core Introduction

H2O Core

Page 8: H2O Core Introduction

H2O Core

H2O

Page 9: H2O Core Introduction

H2O Core

CPU

Page 10: H2O Core Introduction

H2O Core

CPU

Page 11: H2O Core Introduction

H2O Core

CPU

Model Building

Page 12: H2O Core Introduction

H2O Core

H2O

H2O

H2O

Page 13: H2O Core Introduction

H2O Core

CPU CPU CPU

Page 14: H2O Core Introduction

H2O Core

CPU CPU CPU

Model Building

H2O Distributed In-Memory

Page 15: H2O Core Introduction

H2O Core

YARN

CPU CPU CPU

Page 16: H2O Core Introduction

H2O Core

YARN

CPU CPU CPU

Model Building

H2O Distributed In-Memory

Page 17: H2O Core Introduction

H2O Core

YARN

CPU CPU CPU

Model Building

H2O Distributed In-Memory

SQL NFS

S3

Page 18: H2O Core Introduction

H2O Core

YARN

CPU CPU CPU

Model Building

H2O Distributed In-Memory

SQL NFS

S3

Models

Binary

MOJO

POJO

Page 19: H2O Core Introduction

H2O Cluster

H2O

Page 20: H2O Core Introduction

H2O Clients

H2O Cluster

H2O

H2O

Page 21: H2O Core Introduction

H2O Clients

H2O Cluster

REST/

JSON

LocalMachine

H2O

H2O

Page 22: H2O Core Introduction

H2O Clients

H2O Cluster

REST/

JSON

LocalMachine

H2O

H2O

Page 23: H2O Core Introduction

H2O Clients

H2O Cluster

REST/

JSON

LocalMachine

H2O

Page 24: H2O Core Introduction

H2O Clients

JVM 1

JVM 2

JVM N

REST/

JSON

LocalMachine

H2O

Page 25: H2O Core Introduction

H2O Clients

JVM 1

JVM 2

JVM N

H2O Cluster

REST/

JSON

LocalMachine

H2O

Page 26: H2O Core Introduction

H2O Clients

JVM 1

JVM 2

JVM N

x1 x2 x3 xp y

H2O Cluster

REST/

JSON

LocalMachine

H2O

Page 27: H2O Core Introduction

H2O Clients

JVM 1

JVM 2

JVM N

x1 x2 x3 xp y

H2O Cluster

REST/

JSON

LocalMachine

H2O

Page 28: H2O Core Introduction

Current Algori thm Overview

Statistical Analysis

• Linear Models (GLM)

• Naïve Bayes

Ensembles

• Random Forest

• Distributed Trees

• Gradient Boosting Machine

• Stacking / Super Learner

Deep Neural Networks

• MLP

• Autoencoder

• Anomaly Detection

• Deep Features

• CNN, RNN (Deep Water)

Clustering

• K-Means (Auto-K)

Dimension Reduction

• Principal Component Analysis

• Generalized Low Rank Models

Word Embedding

• Word2Vec

Time Series

• iSAX

Machine Learning Tuning

• Hyperparameter Search

• Early Stopping

Page 29: H2O Core Introduction

H2O Flow

Page 30: H2O Core Introduction

H2O R Interface

Page 31: H2O Core Introduction

H2O Python Interface

Page 32: H2O Core Introduction

Deployment Code

YARN

CPU CPU CPU

Model Building

H2O Distributed In-Memory

SQL NFS

S3

Models

Page 33: H2O Core Introduction

Deployment Code: Plain Old Java Object (POJO)

POJO

Page 34: H2O Core Introduction

Current Algori thm Overview

Statistical Analysis

• Linear Models (GLM)

• Naïve Bayes

Ensembles

• Random Forest

• Distributed Trees

• Gradient Boosting Machine

• R Package - Stacking / Super

Learner

Deep Neural Networks

• Multi-layer Feed-Forward Neural

Network

• Auto-encoder

• Anomaly Detection

Clustering• K-Means

Dimension Reduction

• Principal Component Analysis

• Generalized Low Rank Models

Solvers & Optimization

• Generalized ADMM Solver

• L-BFGS (Quasi Newton Method)

• Ordinary Least-Square Solver

• Stochastic Gradient Descent

Data Munging

• Scalable Data Frames

• Sort, Slice, Log Transform

• Data.table (1B rows groupBy record)

Text Processing

• Word2Vec

Page 35: H2O Core Introduction

What is new – Driverless AI

• https://techcrunch.com/2017/07/06/h2o-ais-driverless-ai-automates-machine-learning-for-businesses/

Page 36: H2O Core Introduction

Helpful resources

• Docs

o http://docs.h2o.ai/h2o/latest-stable/index.html

• H2O User Guide

o http://docs.h2o.ai/h2o/latest-stable/h2o-docs/index.html

• Source Code

o https://github.com/h2oai/

o https://github.com/h2oai/h2o-3

o https://github.com/h2oai/sparkling-water

o https://github.com/h2oai/deepwater

• Meetup content

o https://github.com/h2oai/h2o-meetups

• Tutorials

o https://github.com/h2oai/h2o-tutorials

Page 37: H2O Core Introduction

Thank you so much!!

ありがとうございました


Recommended