Dataiku for Enterprise AI and Machine Learning ( dataikudss | 21 hours )

• Experience with Python, SQL, and R programming languages
• Basic knowledge of data processing with Apache Hadoop and Spark
• Comprehension of machine learning concepts and data models
• Background in statistical analyses and data science concepts
• Experience with visualizing and communicating data

Audience

• Engineers
• Data Scientists
• Data Analysts

Dataiku Data Science Studio (Dataiku DSS) is a centralized platform for operating and implementing machine learning models in enterprise applications. It also allows its users to collaborate and iterate on various AI and ML approaches using computation abstraction features. Dataiku DSS, along with its supported instances on different cloud services, provides a solution to big data management that is aligned with the goals of the business.

This instructor-led, live training (online or onsite) is aimed at engineers, data scientists, and data analysts who wish to use Dataiku DSS for the development of machine learning pipelines, and leverage AI processes in generating strategic organizational initiatives.

By the end of this training, participants will be able to:

• Install and configure Dataiku DSS upon their preferred operating system.
• Understand AI/ML concepts and principles that are fundamental to Dataiku features.
• Create and implement their own data visualization codes in Dataiku DSS projects.
• Deploy ML models and pipelines into production environments built around Dataiku DSS.
• Optimize the agility and flexibility of data analysis methods of their enterprise applications.
• Utilize Dataiku DSS to secure and manage how data flows through further business systems.

Format of the Course

• Interactive lecture and discussion.
• Lots of exercises and practice.
• Hands-on implementation in a live-lab environment.

Course Customization Options

• To request a customized training for this course, please contact us to arrange.

Introduction

Installing and Configuring Dataiku Data Science Studio (DSS)

• System requirements for Dataiku DSS
• Setting Up Apache Hadoop and Apache Spark integrations
• Configuring Dataiku DSS with web proxies
• Migrating from other platforms to Dataiku DSS

Overview of Dataiku DSS Features and Architecture

• Core objects and graphs foundational to Dataiku DSS
• What is a recipe in Dataiku DSS?
• Types of datasets supported by Dataiku DSS

Creating a Dataiku DSS Project

Defining Datasets to Connect to Data Resources in Dataiku DSS

• Working with DSS connectors and file formats
• Standard DSS formats v.s. Hadoop-specific formats
• Uploading Files for a Dataiku DSS Project

Overview of the Server Filesystem in Dataiku DSS

Creating and Using Managed Folders

• Dataiku DSS recipe for merge folder
• Local v.s. non-local managed folders

Constructing a Filesystem Dataset Using Managed Folder Contents

• Performing cleanups with a DSS code recipe

Working with Metrics Dataset and Internal Stats Dataset

Implementing the DSS Download Recipe for HTTP Dataset

Relocating SQL Datasets and HDFS Datasets Using DSS

Ordering Datasets in Dataiku DSS

• Writer ordering v.s. read-time ordering

Exploring and Preparing Data Visuals for a Dataiku DSS Project

Overview of Dataiku Schemas, Storage Types, and Meanings

Performing Data Cleansing, Normalization, and Enrichment Scripts in Dataiku DSS

Working with Dataiku DSS Charts Interface and Types of Visual Aggregations

Utilizing the Interactive Statistics Feature of DSS

• Univariate analysis v.s. bivariate analysis
• Making use of the Principal Component Analysis (PCA) DSS tool

Overview of Machine Learning with Dataiku DSS

• Supervised ML v.s. unsupervised ML
• References for DSS ML Algorithms and features handling
• Deep Learning with Dataiku DSS

Overview of the Flow Derived from DSS Datasets and Recipes

Transforming Existing Datasets in DSS with Visual Recipes

Utilizing DSS Recipes Based on User-Defined Code

Optimizing Code Exploration and Experimentation with DSS Code Notebooks

Writing Advanced DSS Visualizations and Custom Frontend Features with Webapps

Working with Dataiku DSS Code Reports Feature

Sharing Data Project Elements and Familiarizing with the DSS Dashboard

Designing and Packaging a Dataiku DSS Project as a Reusable Application

Overview of Advanced Methods in Dataiku DSS

• Implementing optimized datasets partitioning using DSS
• Executing specific DSS processing parts through computations in Kubernetes containers

Overview of Collaboration and Version Control in Dataiku DSS

Implementing Automation Scenarios, Metrics, and Checks for DSS Project Testing

Deploying and Updating a Project with the DSS Automation Node and Bundles

Working with Real-Time APIs in Dataiku DSS

• Additional APIs and Rest APIs in DSS

Analyzing and Forecasting Dataiku DSS Time Series

Securing a Project in Dataiku DSS

• Managing Project Permissions and Dashboard Authorizations
• Implementing Advanced Security Options

Integrating Dataiku DSS with The Cloud

Troubleshooting

Summary and Conclusion