Productionizing Machine Learning with Delta Lake
Try out this notebook series in Databricks - part 1 (Delta Lake), part 2 (Delta Lake + ML) For many data scientists, the process of building and tuning machine learning models is only a small portion of the work they do every day. The vast majority of their time is spent doing the less-than-glamorous (but...
Announcing Databricks Runtime 5.5 with Conda (Beta)
Databricks is pleased to announce the release of Databricks Runtime 5.5 with Conda (Beta). We introduced Databricks Runtime 5.4 with Conda (Beta), with the goal of making Python library and environment management very easy. This release includes several important improvements and bug fixes as noted in the latest release notes [Azure|AWS]. We recommend all users...
Automated Hyperparameter Tuning, Scaling and Tracking: On-Demand Webinar and FAQs now available!
On June 20th, our team hosted a live webinar—Automated Hyperparameter Tuning, Scaling and Tracking on Databricks—with Joseph Bradley, Software Engineer, and Yifan Cao, Senior Product Manager at Databricks. Automated Machine Learning (AutoML) has received significant interest recently because of its ability to shorten time-to-value for data science teams and maximize the predictive performance of models....
Scaling Genomic Workflows with Spark SQL BGEN and VCF Readers
In the past decade, the amount of available genomic data has exploded as the price of genome sequencing has dropped. Researchers are now able to scan for associations between genetic variation and diseases across cohorts of hundreds of thousands of individuals from projects such as the UK Biobank. These analyses will lead to a deeper...
Accurately Building Genomic Cohorts at Scale with Delta Lake and Spark SQL
This is the second post in our “Genomic Analysis at Scale” series. In our first post, we explored a simple problem: how to provide real-time aggregates when sequencing large volumes of genomes. We solved this problem by using Delta Lake and a streaming pipeline built using Spark SQL. In this blog, we focus on the more advanced...
Hyperparameter Tuning with MLflow, Apache Spark MLlib and Hyperopt
Hyperparameter tuning is a common technique to optimize machine learning models based on hyperparameters, or configurations that are not learned during model training. Tuning these configurations can dramatically improve model performance. However, hyperparameter tuning can be computationally expensive, slow, and unintuitive even for experts. Databricks Runtime 5.4 and 5.4 ML (Azure | AWS) introduce new...
Announcing the MLflow 1.0 Release
MLflow is an open source platform to help manage the complete machine learning lifecycle. With MLflow, data scientists can track and share experiments locally (on a laptop) or remotely (in the cloud), package and share models across frameworks, and deploy models virtually anywhere. Today we are excited to announce the release of MLflow 1.0. Since...
Enhanced Hyperparameter Tuning and Optimized AWS Storage with Databricks Runtime 5.4 ML
We are excited to announce the release of Databricks Runtime 5.4 ML (Azure | AWS). This release includes two Public Preview features to improve data science productivity, optimized storage in AWS for developing distributed applications, and a number of Python library upgrades. To get started, you simply select the Databricks Runtime 5.4 ML from the...
Introducing Databricks Runtime 5.4 with Conda (Beta)
We are excited to introduce a new runtime: Databricks Runtime 5.4 with Conda (Beta). This runtime uses Conda to manage Python libraries and environments. Many of our Python users prefer to manage their Python environments and libraries with Conda, which quickly is emerging as a standard. Conda takes a holistic approach to package management by...
Advanced Analytics with HyperLogLog Functions in Apache Spark
This is a community guest blog from Sim Simeonov, the founder & CTO of Swoop and IPM.ai. Pre-aggregation is a common technique in the high-performance analytics toolbox. For example, 10 billion rows of website visitation data per hour may be reducible to 10 million rows of visit counts, aggregated by the superset of dimensions used...
