Speculative Life member, Gabriel Vigliensoni leads a workshop aimed at presenting the goals of machine learning, understanding the basic rationale of optimization algorithms and models, and the current approaches and frameworks. This is the first series of workshops lead by members of Speculative Life.
Machine learning applications have penetrated our daily experience and are ubiquitous. We find travel deals, do online shopping, and search for recommendations without thinking about how “the machine” learned something about us. What we write about in our emails, what we listen to or watch in streaming services, the pictures we post, our location, and our demographic characteristics are data that are collected, filtered, and aggregated in order to create models of ourselves.
In the first half of this workshop we will aim at presenting the goals of machine learning, understanding the basic rationale of optimization algorithms and models, and the current approaches and frameworks. We will also review some of the most common datasets and tasks for benchmark, as well as looking at some derivative work that has been seen as the “artistic output” that some of these optimization algorithms and models can produce. Wekinator, a free, open source machine learning software aimed at learning input signals (such as human gestures) and map them to any real-time process (such as trigger events or modulate other signals), will be presented in the second half of the workshop. With this software, non-coders will be able to create models to map gestures to some output by providing training examples from which the learning algorithm of the application can learn. [Watch some examples of artwork using Wekinator here].
April 5 at 2pm-6pm
Speculative Life Space, EV.10.625
Concordia University SGW,
1515 Saint Catherine Street W.
This workshop is open to everyone (with priority to Speculative Life members). Attendance is free but limited, registration required.
Gabriel Vigliensoni (PhD in Music Technology, McGill Univeristy) has been professionally involved in popular music production, performance, and sound recording for several artists and record labels for the past 20 years.