Steven Dorsher

Former Computational Physicist

GitHub Home Page

ORCID

GitHub

ResearchGate

I have had a 15+ year research career in physics and astronomy over the course of my bachelors degree, three masters degrees, and since my final graduation. Most of this (10+ years) has been in computational physics. My primary focus is in general relativity, black holes, and gravitational waves. I have also done research in cosmology, particle physics, neutrinos, exoplanets, the three body Newtonian gravity problem, and fractional calculus.

I have 8 main publications/documents and about 35 total including membership in LIGO spanning the time of the first three detections.

My background is primarily in scientific programming, in the context of data analysis algorithms, experiment development and assessment, and numerical methods for theory.

Education:

MS Physics, LSU, December 2017
MS Physics, U of MN, July 2013
MS Astronomy, Ohio State, August 2006
BS Physics, MIT, June 2004

Strongest Physics Fields:

General Relativity
Black Holes (EMRI's, LISA)
Gravitational Waves (LIGO)
Particle Physics

Other Fields

Neutrinos (MINOS, NOvA)
Exoplanets
Cosmology
Fractional Calculus

Scientific Computing Languages:

Currently most fluent in:
- Python
Strong previous experience with:
- C++/C (6.5 years)
- Fortran (5 years)
Other previous experience with:
- Matlab (2.5 years)
- Java (1 year)
Technically have used:
- Bash
- Scheme
- Haskell

Machine Learning-- beginner level, certificates, need practical experience still, mostly guided projects so far

Data Cleaning
Visualization
Regression-- linear, relu, logistic, tanh, softmax
Classification
Clustering
Neural Networks
Convolutional Filtering
Adam optimizer
Batch filter
Recurrent Neural Networks
Maybe: Transformers
Maybe: GRUs, LTSMs
Guided Projects in: phrase generation, image recognition and object detection and location, music generation, activation word detection (and overlay of beep), facial recognition and authorization, video game starship lander trajectory prediction, French to English translation (though that was very heavily guided and my French is very beginner)
Mathematical calculation of more precise technique for cosine similarity (it is possible a reference was given but I computed it instead)

Databases

Pandas
A little SQL
A little HBase

Although very rusty, past experience with:

Linux
MacOS
Windows
Gnuplot
LaTeX
Valgrind (a little, and rusty)
gdb (a little, and rusty)
github
svn (a little, and very very rusty)
OpenMP (a little, and rusty)
MPI (a little, and rusty)
mpi4py (a little, and rusty)

Languages Spoken

ENGLISH--native (Minnesota... dialect shows)
Danish-- I've finished the Duolingo track but I have little practical speaking experience and my reading and writing are not outstanding either.
French-- Definitely high school from 2000 level, although I've returned to it on Duolingo and completed the track
I have learned extremely small amounts of ASL, German, Japanese, Turkish, and Spanish. If you live anywhere south of central Minnesota, it will shock you how bad my Spanish is. My hearing is not good, and I use visual cues heavily and you will also be shocked by what I can't hear, but I am not Deaf and do not have an interpreter or hearing aid or cochlear implant. Please be skeptical that we have the same information or are even hearing the same thing.

On names and ID: My degrees were granted under the name of Susan Dorsher, which is my legal name (I have sent LinkedIn ID in the form of my passport card and MN state ID). The majority of my publications are under Steven Dorsher (please see ORCID or ResearchGate where both names are listed) to reflect the name I have used personally and professionally since 2009. Please, I go by Steven.

This is my real name and identity. I do not have alternate identities. My Guild Wars 2: WaveIn; SWTOR legacy: Dreamstar. FFXIV: Eternal Spiralduality. MIT & 90's: eirl; Legend's LARPing in 2000's: Aurbren

Page updated

Google Sites

Report abuse