Colloquium Talks

To avoid repetitive analysis, environmental impact factors (EIF) of common materials and products are published for use by LCA experts. However, finding appropriate EIFs for even a single product under study can require hundreds of hours of manual work, especially for complex products. We present Flamingo, an algorithm that leverages natural language machine learning (ML) models to automatically identify an appropriate EIF given a text description. 

When a skilled machine learning engineer is tasked with building a system for a specific application, they take several steps. Some of these include doing a literature review of the most appropriate models and datasets, choosing which ones to utilize based on accuracy and other constraints such as efficiency or latency, creating or curating training and testing data, training and comparing models, identifying weak points of the current modeling paradigm and iteratively improving.

 In this talk, I will focus on recent projects that explore how technologies can facilitate empathy. These approaches primarily focus on those who need to be empathized and help them express, reveal, and reflect on themselves. 

In this talk, I'll present a new perspective on the data representation substrate for data platforms. I argue that a bit-level data shredding approach could offer significant advantages for future data platforms. 

In this talk, the speaker will share his experience on conducting experimental research. In particular, he will talk about three backscatter sensing and communication systems he built in the past ten years and share the lessons he learnt from building them, as well as discussing how he leverages these lessons to expand his research.

We develop an algorithm for parameter-free stochastic convex optimization (SCO) whose rate of convergence is only a double-logarithmic factor larger than the optimal rate for the corresponding known-parameter setting.