About
Hi! I'm Andrew Portal, a Data Scientist. I utilize programming languages to tie Machine Learning and branches of mathematics together into signal processing applications. In case the significance of this is unclear, consider the simple question of how many items are in a given room. This may seem trivial, but if you think critically you run into the fundamental problem of how to best differentiate objects. Each object can be further broken down into smaller objects across an infinite set of dimensions. Determining the best intersection of dimensions and the ultimate count objects is signal processing. The human mind naturally processes signals from our environment into discernable sights and sounds, so the process of counting objects seems trivial but becomes profound when we try to program it and apply it to dimensions of analysis beyond just sight and sound. To expand the domain of human knowledge and capability, the development of signal processing application is essential.
Grit & Grid?
In machine learning, grid search refers to cycling hyper-parameters of a model to find the optimal set. Grit comes into play because inevitably some exhaustive manual process of trial and error comes into play when developing. Together these words articulate the process of I building machine learning applications … and happen to form an alliteration.
How I got started?
While working as an analyst for several years after graduating, I began programming to automate workflows. By working with promgraiming languages, I realized the greater utilty of computers. In the Summer of 2018 I moved to San Francisco to pursue a more technical career. I spent the rest of the year studying programming and attending tech meetups in the area, where I encountered machine learning.
With machine learning, I can use programming to not only automate tasks but also enable feats previously impossible. For example, I met a group using EEG information along with ML to detect emotional states and then build a VR environment to respond. Rather than just automate a process a human can do manually, ML enablaled a new frontier of possibilities. After discovering this, I knew where I needed to focus my time.
I enrolled in Metis, a data science training program I completed Jan-Mar 2019. Along the way I’ve learned a lot and compiled several projects, which I have tried to describe and share in this blog.
Other places you can find me on the web:
Thanks for reading!