Snowflake Data Science Training

Course Outline

Snowflake Data Cloud Architecture and Overview

• Snowflake Data Cloud overview
• Three-tiered architecture
• Snowflake UI and core capabilities, including elasticity, workload separation, data security and simplicity
  of performance

Data Exchange and Data Marketplace

• Private and Public Data Exchange
• Data Marketplace with ready-to-use and third-party datasets for data augmentation
• Diverse data including customer demographic data, time-series data, geospatial data
• Exploration and visualization using Snowsight

Data Lake for Machine Learning and Analytics

• Raw and external data sets in object stores
• External tables and direct queries in data lakes
• Native data formats of such as CSV, JSON, Parquet

Data Ingestion service and Continuous Data Pipelines

• Serverless continuous ingestion service Snowpipe
• Data ingestion best practices
• Bulk ingestion and scheduling data loads with tasks
• Table stream for capturing change data

Working with Semi-Structured Data

• Ingesting into native semi-structured data types without pre-processing
• Built-in functions for traversing, flattening, and nesting of semi-structured data
• Leveraging semi-structured state data for JavaScript Stored Procedures
• Complement learning with topics like geospatial data

Data Science and Machine Learning Concepts and Applications

• Data science applications
• Common machine learning vocabulary
• Machine learning workflow and pipeline
• Supervised and unsupervised machine learning

Data Science Toolset and Ecosystem

• Seamless connectivity using Snowflake connectors for languages such as Python, Spark, and R
• Notebook-based data science development environments
• Open source and many machine learning libraries including Scikit-Learn and more
• Partner platforms for data science automation and democratization around AutoML
• Partner platforms for deployment and practices with MLOps

Exploratory Data Analysis and Feature Engineering

• Descriptive exploratory data analysis using statistical and analytic functions
• Visual exploratory data analysis using popular and relevant libraries
• Employ common feature selection and feature engineering techniques
• Advanced SQL functions for data transformation at scale

Machine Learning Model Development and Tuning

• Supervised learning: linear regression with popular ML libraries
• Supervised learning: classification using techniques such as logistic regression, random forests, gradient boosts and more
• Identifying, using, and interpreting metrics to evaluate models and performance
• Unsupervised learning

Model Management and Deployment at Scale

• Deploying machine learning models using scalable framework
• External functions to support prediction and data augmentation through APIs
• Extensive partner ecosystem for automation around AutoML and operationalization using MLOps practices
• Using Snowflake capabilities including Snowpipe, table stream, and tasks for continuous data pipelines to update machine learning models
• Storing machine learning results in Snowflake

Visualizing and Collaborating on Data Science and Machine Learning Results

• Seamless connectivity to BI tools for reporting and analytics
• Communicating machine learning results
• Collaborating on models by sharing results with data sharing techniques
• Replicating your raw and processed data across region and cloud providers including AWS, Azure, and GCP