The idea of a universal approximator has been around since the early 1900s. The mathematical proofs required to prove that you can transform any function into a set of inner and outer functions (approximation) was shown by Andrey Kolmogorov and Vladimir Arnold (his student) in the early 60s. It was called the Kolmogorov-Arnold Representation Theory and was later used to develop what we now know as “Deep Learning”. In Academia Letters, I published a letter that draws a a comparison […]
Abstract The firm size distribution describes important economic and labor properties of any economy. Government entities must expend enormous resources in data collection, cleaning, and analysis in order to construct this and other important distributions describing the aggregate properties large economies. In the U.S., this process can be cumbersome and relies on querying multiple databases and utilizing significant computational resources. I show that construction of the U.S. firm size distribution is plausible using only individual income tax records (W2s) drawn […]
Here’s the link: https://dl.acm.org/doi/10.5555/3213032.3213048
We developed an agent-based model in order to understand agent/node behaviors that generate social media networks. We use simple rules to synthetically generate a backcloth (friend/follow) network collected using Twitter’s API. The Twitter network was collected using seeds for known terrorist propaganda accounts in 2015. Model parameter adjustments were made to reproduce the collected network’s summary statistics, stylized facts and general structural measures. We produced an approximate network in line with the general properties of our collected data. We present our findings with a focus on the challenging aspects of this reproduction. We find that while it is possible to generate a social media network utilizing a few simple rules, numerous challenges arise requiring departure from the agent viewpoint and the development of more useful methods. We present numerous weaknesses and challenges in our reproduction and propose potential solutions for future efforts.