machine learning in the cloud: building a better forecast with h20 & salesforce
TRANSCRIPT
Machine Learning in the Cloud
Mark Masterson Application Engineer [email protected] @MarkMastersonSF
Building a better Forecast with H2O and Salesforce
Basics of Machine Learning
H2O and how it can be used to perform Machine Learning tasks
How Salesforce can be enhanced by integrating with Machine Learning tools
Agenda
Machine Learning consists of building predictive models that are learned from data.
Supervised Learning: given data x, predict y
• Two major uses: classification and regression
Unsupervised Learning: What are interesting things about data x
Basics of Machine Learning
H2O - http://h2o.ai/
Amazon Machine Learning - https://aws.amazon.com/machine-learning/
Microsoft Azure Machine Learning - http://azure.microsoft.com/en-us/services/machine-learning/
Google Prediction API - https://cloud.google.com/prediction/docs
Apache Spark - http://spark.apache.org/
Machine Learning Offerings
Open Source predictive analytics platform
Easy to use Web Interface
NanoFast™ Scoring Engine
Robust REST API Support
Distributed and In-Memory
Support for Java, Python, R, and Scala
Why H2O?
Deep Learning
Distributed Random Forest
Gradient Boosting Machine (GBM)
Generalized Linear Model (GLM)
K-Means
Native Bayes
H2O Algorithms
Advantages
• Gives sales users a quick view into the current status, and values of opportunities
• Excellent User Interface for Sales Managers and Sales People to collaborate on sales goals
Disadvantages
• Not a true “forecast” informed by statistical analysis
• Relies on human input and analysis, which is prone to error.
Forecast Advantages and Disadvantages
Data supplemented with prediction models provided by H2O
Integration with H2O gives us access to highly performant Machine Learning
Solution: Forecasting 2.0
Resources
• H2O – http://h2o.ai/product/
• H2O Github – https://github.com/h2oai/h2o-3/
Questions and Answers Q&A