Virtual Meetup: Event Data Processing and Model Performance

May 11, 09:00 AM PDT
  • Virtual AICamp
  • 136 RSVP

Welcome to the weekly AI/ML/Data virtual meetup, which you can join from anywhere around the world. Join us for deep dive tech talks on AI/ML/Data, networking with speakers&peer developers, and win lucky draw prizes.

Agenda (PDT):
* 9:00am: Welcome/community update/Sponsor intro
* 9:10am: Tech talks
* 10:30am: Opening discussion, Lucky draw and Mixer

Tech Talk: Declarative Reasoning with Timelines: The Next Step in Event Processing
Speaker: Ryan Michael, VP of Engineering @Datastax
Abstract: At the heart of modern data processing lies events. Events describe the roughest, most complete picture available of what has happened in the world, and practically every form of data processing ultimately begins with events.
While the power of event processing has increased since the emergence of streaming data processing, current systems are still difficult to use when working on problems that deal with time and order, such as predictive AI/ML. Handling these problems requires a new kind of query language - a way to declaratively reason about events over time.
In this talk, we introduce the concept of timelines. Timelines are an intuitive abstraction for reasoning about temporal values. They support a broad range of useful operations which can be efficiently computed at scale. We will demonstrate the power and differentiation of timelines:
- How timelines allow declarative queries over events and time in a simple and intuitive manner
- Why timelines are ideal for applications such as behavioral predictions, trend analysis, and forecasting, and how existing solutions such as streaming SQL fall short.
- How to execute timeline based queries using the open-source Kaskada event-processing engine.

Tech Talk 2: Turbocharging Model Performance by Super-sizing Your Data
Speaker: Ram Seshadri, Senior Program Manager @Google
Abstract: As data scientists and ML engineers, we often dedicate countless hours to fine-tuning hyperparameters of complex algorithms, particularly neural networks, in pursuit of optimal model performance. This talk will introduce a set of powerful open-source libraries, developed by the speaker, that enable a similar level of optimization for your data itself. Discover the capabilities of pandas-dq, featurewiz, Sulo model class, and lazytransform libraries as they empower you to enhance data quality, augment your data through feature engineering, and streamline feature selection for tabular datasets, including time series data.
The presentation will include a detailed walkthrough of an example notebook, illustrating how these innovative libraries can be harnessed to boost model performance in tabular and time series data contexts. We will explore strategies for refining data quality, addressing target imbalances, optimizing cross-validation, and effectively executing feature engineering and feature selection. Elevate your models to new heights by unlocking the full potential of your data, and learn how to troubleshoot underperforming segments for even greater results.

Lucky draw
We will raffle winners for prizes during the event. To enter the lucky draw, share the event on social media:

  • Prefer Twitter? Twitter the event with hashtag #aicampvirtual and tag @aicampai. For example:
  • #aicampvirtual Join the virtual ML meetup by @aicampai to learn AI, ML, Data and Cloud technology with tech leads and industry experts. Free join in person:
  • Prefer LinkedIn? Comment the post on LinkedIn: LinkedIn Post
  • Community on Slack
    - Event chat: chat and connect with speakers and attendees
    - Sharing blogs, events, job openings, projects collaborations
    Join Slack (search and join the #virtualevents channel)

    Ryan Michael (DataStax),Ram Seshadri (Google)

    Ryan Michael
    Ryan leads strategy and community development for Kaskada, an open-source project focused on a novel event processing engine that utilizes stream-native query abstractions to simplify time-based reasoning. As Kaskada VP of Engineering and technology co-creator, Ryan was instrumental in the company evolution from a SaaS startup to an open-source project, following its acquisition by DataStax earlier this year. Before Kaskada, Ryan served as a principal engineer at ReturnPath.
    Ram Seshadri
    Senior Program Manager at Google. Ram is a Data Scientist with deep expertise and experience in financial services. He is also a popular speaker and creator of several open source libraries including AutoViML and AutoViz. Prior to ML, Ram was a trader and portfolio manager on Wall Street for more than 10 years. He also worked as a Data Scientist at Morgan Stanley and an instructor for General Assembly & the New York Institute of Finance.
    The event ended.
    Watch Recording
    *Recordings hosted on Youtube, click the link will open the Youtube page.
    Contact Organizer