Return to page

Make with

Free Weekly Sessions | 8am PT

Register Now Watch On-Demand Videos

The Make with webcast program is designed to educate data scientists, data engineers, analysts, and developers on the latest in data science techniques, technologies, and successful use cases.  The objective of the program is to address the needs of the data science and AI community to help them make better AI solutions that solve real world problems.

All of the sessions are complimentary, and we will run a session just about every week on Tuesdays at 8am PT.  Occasionally, there will be a holiday or a bigger event, but for the most part, we’ll host them every week.  We’ll always have the next four upcoming sessions on the page, just register for the ones that are the most interesting to you.

Sessions will be led by product managers, engineers, Kaggle Grandmaster, data science luminaries, and customers.  We’ll continue to poll the community to hear what you want to learn about most, so please provide input on our social polls as well as polls within the webcasts.

Upcoming Sessions

August 2022

Title Title
North America: August 9th, 2022 | 8am PT
August 11th, 2022 | 1PM SGT

Accuracy Masterclass Part 2 - Validation Scheme Best Practices

Setting up a validation strategy is one of the most crucial steps in creating a machine learning model. A poorly designed validation scheme can lead to a major model accuracy overestimation or even completely erroneous model. There are common concepts of how to set up a validation strategy and what are typical mistakes to avoid. 

3 Main Learning Points

• Learn why it is important to do validation for machine learning

• Overview the common validation methods

• See examples of good and bad validation approaches

Eric Gudgion Eric Gudgion

Dmitry Gordeev
Kaggle Grandmaster

Title Title
North America: August 16th, 2022 | 8am PT
August 18th, 2022 | 1PM SGT

Accuracy Masterclass Part 3 - Feature Selection Best Practices

Not all features are created equal. It is tempting to put all available features into the model but if you have too many features or features which are unrelated and/or noisy this could hurt model's performance. Finding best feature subset could be very useful in industrial application: • it enables model to train faster • reduces complexity of data pipeline • reduces overfitting • might improve performance. Sometimes, less is better.

3 Main Learning Points

  • Learn when to use feature selection and why

  • Understand main approaches to feature selection and their tradeoffs

  • See examples of feature selection techniques
Eric Gudgion Eric Gudgion

Dmitry Larko
Kaggle Grandmaster

Title Title
North America: August 23rd, 2022 | 8am PT
August 25th, 2022 | 1PM SGT

Getting Started with H2O Driverless AI

H2O Driverless AI is a fully customizable award-winning AutoML platform that empowers data scientists to work on projects faster and more efficiently. It automates data preparation, feature engineering, model validation, model tuning, model selection and model ensembling, and also provides scoring pipelines for rapid standalone deployment out of the box, as well as model interpretability. In this webinar we'll demonstrate a new graphical wizard that makes it easy to create highly accurate models for any tabular dataset, including time series, image and text use cases.

3 Main Learning Points

  • Learn about the capabilities of H2O Driverless AI as the AutoML platform of choice for tabular datasets

  • Learn how to build sophisticated models in minutes using the new graphical Wizard

  • See examples of use cases that are ideal for H2O Driverless AI


Eric Gudgion Eric Gudgion

Arno Candel
Chief Technology Officer

Title Title
North America: August 30th, 2022 | 8am PT

Accuracy Masterclass Part 4 - Time Series Modeling

This webinar will dive deeper into workflows for time series modeling such as forecasting. We'll show how to make sure that temporal causality is preserved during modeling, how to automatically generated lag-based features, how to deal with trends, and how to measure the performance of the model as time advances.

3 Main Learning Points

  • Learn how to solve time-series forecasting problems with H2O Driverless AI

  • Learn how to backtest and validate models

  • Learn how to deploy time-series models in production
Eric Gudgion Eric Gudgion

Megan Kurka
Product Manager

Register Now

Please select the time zone you would like to attend in.

Make with H2O will be offered live in real time in North America. In Asia-Pacific, we will rerun the broadcast live with moderators to answer questions.

Catch up and Watch On Demand Session Videos