Data Science Friday Webinars
About This Webinar Series
Join us each Friday to learn about leveraging the power of the Wolfram Language for data science. In the first collection of this multipart series, we will focus on the wide range of state-of-the-art integrated machine learning capabilities available in the Wolfram Language. We'll start with the first steps of importing data from many different sources and getting it ready for machine learning. Then we will look at highly automated functions like Predict, Classify, AnomalyDetection and FeatureExtraction, as well as the powerful symbolic Wolfram Neural Net Framework. The series will also include pointers to freely available repositories of neural net models and curated computable datasets, so you can jump right in and start your data science explorations. Each webinar is led by a Wolfram certified instructor and content expert who will cover weekly topics, poll the audience about the addition of new topics for the series and answer your questions.
Webinars in This Series
  • Friday, March 25, 1–2pm US CDT (6–7pm GMT)
    Getting Your Data Ready for Automated Machine Learning
    Learn about the many different import and export formats available in the Wolfram Language. See how easy it is to use the built-in Wolfram Knowledgebase and to access the curated datasets in the Wolfram Data Repository. We'll also look at getting data from local files, from the web and from a database. We will discuss restructuring data and extracting parts of data. The webinar ends with a few examples demonstrating how the restructured data can be used for further analysis and visualization.
  • Friday, April 1, 1–2pm US CDT (6–7pm GMT)
    Machine Learning and Statistics: Better Together
    Join an interactive and example-driven exploration showcasing the computational capabilities of the Wolfram Language in the fields of machine learning and classical statistics. This presentation demonstrates how the powerful symbolic nature of the Wolfram Language makes the handling of statistical distributions simple, how automation can play a part in making machine learning accessible and how the two fields together can allow the utilization of some powerful and flexible tools.
  • Friday, April 8, 1–2pm US CDT (6–7pm GMT)
    Supervised and Unsupervised Learning
    Learn how basic supervised and unsupervised learning tasks can be accomplished with the Wolfram Language. This webinar will show you how to use automated superfunctions like Predict and Classify to build your own models from labeled training data. In the absence of labeled data, functions like LearnDistribution and FindClusters can be used to automatically discover underlying structures in your data. FindAnomalies and DeleteAnomalies can be used to detect anomalies and remove them from the data. Functions like FeatureExtraction, DimensionReduction and FeatureSpacePlot, which provide great tools to explore your data in the feature space, will also be covered.
  • Friday, April 22, 1–2pm US CDT (6–7pm GMT)
    Human in the Loop: Interpretable Machine Learning
    As machine learning tools become more powerful, the importance of interpreting them is more important than ever. Understanding why your model gives its answer is critical in trusting it, diagnosing problems and making improvements. Computers are efficient at spotting patterns, but with a human in the loop, we can interpret the patterns better. We will highlight tools in the Wolfram Language for gaining insight into your machine learning models.
  • Friday, April 29, 1–2pm US CDT (6–7pm GMT)
    Exploring the Neural Network Framework from Building to Training
    Discover the easy-to-use framework available in the Wolfram Language to build, modify, train and deploy neural networks. Learn how the Wolfram Language simplifies the process of encoding input and decoding output for neural nets. We'll also introduce the many types of layers—the basic building blocks for constructing neural networks—and the process of connecting them in chains and graphs to build complicated networks according to your requirements.
  • Friday, May 6, 1–2pm US CDT (6–7pm GMT)
    Ready-to-Use Datasets and Neural Net Models: Two Useful Wolfram Repositories
    Learn how you can freely access thousands of computable curated datasets and neural network models. The Wolfram Data Repository is a system for publishing data and making it available in the Wolfram Language for immediate computation. The Wolfram Neural Net Repository is a public resource hosting an expanding collection of neural network models in the Wolfram Language. We'll explain the motivation behind the repositories, describe their components and provide examples of how to use them.
Join Current Webinar