Wednesday, May 19, 2021 at 11:00:00 AM Online event
Please register at the link below (Must be registered to attend):
https://www.tibco.com/events/dr-data-science-office-hours
This year we launched a new Dr. Data Science series that explores a real-world data science project end to end. We'll take a comprehensive and detailed look at the work of the data scientist by examining each stage of a complex project, from inception through to deployment. We'll not only cover the classical phases of data science — including exploration, data preparation, feature generation, modeling, testing, and model ops — but also look at what happens when the right data isn't available, avenues of exploration don't yield results, models aren't accurate enough, and requirements change.
With the sort of anomaly detection that's central to this project, one frequent challenge is that the data may well be unlabelled. We can apply various 'bootstrapping' techniques to identify anomalies, and in the last few sessions we've seen how clustering can be used to identify targets of interest. In this session, we take a sort of 'semi-supervised' approach so that experts can iteratively refine predictive models. We show how clusters and ensembles — and humans — can work together to create highly accurate models.
What is Dr. Data Science?
Dr. Data Science is an interactive presentation aimed at helping new and advanced practitioners of data science and machine learning. The program is hosted by experts from our Data Science and Customer Success teams, including veterans from the world of advanced analytics and architects of the TIBCOⓇ Data Science platform. The program consists of two parts:
1. Office Hours: Monthly 30-minute live webinars to help everyone navigate best practices and technologies in the world of data science
- Info Gain: A tidbit of useful information for everyday data science, based on the most popular questions submitted to the TIBCO Community for Data Science
- Feature Sessions: The main presentation, featuring extended topics and unique challenges in advanced analytics and machine learning
2. Regular Tips: Weekly 5-10 minute YouTube videos to help users with day-to-day recommendations for TIBCO Data Science software.
Missed any previous sessions? Catch up on our YouTube channel.
How to get your questions answered?
1. Tweet your question with the hashtag, #DrDataScience.
2. Post your question to TIBCO Community "Answers" section with the hashtag, #DrDataScience.
Speakers:
- Steven Hillion, Senior Director, Data Science, TIBCO Software
- Sweta Kotha, Data Scientist, TIBCO Software
0 Response to "May 19: TIBCO Kansas City User Group - Dr Data Science Office Hours"
Post a Comment