Главная
Study mode:
on
1
Introduction
2
What is a feature store
3
How does a business add value
4
ML Operations
5
Feature Store for ML
6
How do I get data into it
7
How do we write feature pipelines
8
Writing data to the feature store
9
Feature views
10
Python
11
Feature Pipeline
12
Feature Quality
13
RealTime Features
14
Feature View
15
Point in Time Correct Join
16
Preengineered features
17
Incremental training
18
Saving training data
19
Hopsworks properties
20
Hopsworks demo
21
Demo project
22
Data preview
23
Feature group activity
24
Training data sets
25
Intrusion detection example
26
Sample log click logs
27
Models
28
Model Data
29
Connecting to the Feature Store
30
Creating an API Key
31
Accessing the Platform
32
Predicting
Description:
Explore the open-source Hopsworks feature store for Python-centric enterprise data science in this 55-minute webinar from Open Data Science. Learn how to manage features for training and serving models, including native Python support for feature engineering, pipelines, low latency access, online transformation functions, and data validation. Discover a Python-based domain-specific language for joining and retrieving features from data warehouses and lakes without SQL. Follow an end-to-end use case demonstrating the process from raw data to operationalized prediction service. Gain insights into feature store concepts, ML operations, data ingestion, feature pipelines, real-time features, point-in-time correct joins, and incremental training. Witness a live demo of the Hopsworks platform, covering project setup, data preview, feature group activities, training datasets, and an intrusion detection example using log click data.

Hopsworks - The Python-Centric Enterprise Feature Store

Open Data Science
Add to list
0:00 / 0:00