Machine Learning with Oryx 2
Course Overview
Acquire in-depth understanding of the architecture of Oryx 2 and be able to deploy each layer to your advantage. Master the Oryx 2 framework to build enterprise grade ML applications by harnessing the power of underlying lambda architecture to figure out models and update the current models as a function of historical data.
In cloud labs, gain hands-on expertise to on in-built ML algorithms to deploy collaborative filtering, classification, regression and clustering in your application.
Be industry-ready to develop offline large scale models, update and query them in time and integrate the models with the app in the fastest possible time with our machine learning with Oryx 2 training.
At the end of the training, participants will be able to:
- Install, configure and test the cluster to support Oryx 2
- Gain a deep understanding of Lambda Tier Implementation
- Use the machine learning tier implementation through built-in interfaces
- Conduct Collaborative filtering based on Alternating Least Squares
- Cluster datasets based on k-means
- Classify and regression based on random decision forests
- Develop, customize and deploy an Oryx app
- Lead intelligent app development projects that leverage on Hadoop architecture
Pre-requisite
Required: Development Experience in Java 8, Scala 2.11, Apache Hadoop Cluster & Services
Duration
3 days
Course Outline
- Oryx 2 History
- Architecture of Oryx 2 Implementation
- Lambda Tier Implementation
- Batch Layer
- Speed Layer
- Serving Layer
- Data Transport
- ML Tier Implementation
- Installation of Java 8
- Installation of Scala 2.11
- Installation of Apache Hadoop Cluster
- Deployment Architecture
- Enabling Services
- Configuring Kafka
- HDFS & Data Layout
- Handling Failure
- Troubleshooting
- Module Mapping
- Creating an App
- Building an App
- Compiling the word-count example
- Customizing an Oryx App
- Deploying an App
- Deploying the word-count example
- Fetch the data
- Push data into Serving layer
- Verify the results
- Implementation of Collaborative Filtering / Recommendation
- Implementation of Classification/Regression
- Implementation of Clustering