-
Notifications
You must be signed in to change notification settings - Fork 0
/
jZink_Resume.tex
220 lines (166 loc) · 12.5 KB
/
jZink_Resume.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
\documentclass[margin,line, 12pt]{res}
\topmargin=-0.5in % Make letterhead start about 1 inch from top of page
\textheight=10in % text height can be bigger for a longer letter
\oddsidemargin -.5in
\evensidemargin -.5in
\textwidth=6.5in
\itemsep=0in
\parsep=0in
\nofiles
% if using pdflatex:
\setlength{\pdfpagewidth}{\paperwidth}
\setlength{\pdfpageheight}{\paperheight}
\newenvironment{list1}{
\begin{list}{\ding{113}}{%
\setlength{\itemsep}{0in}
\setlength{\parsep}{0in} \setlength{\parskip}{0in}
\setlength{\topsep}{0.05in} \setlength{\partopsep}{0in}
\setlength{\leftmargin}{0.17in}}}{\end{list}}
\newenvironment{list2}{
\begin{list}{$\bullet$}{%
\setlength{\itemsep}{0.04in}
\setlength{\parsep}{0.00in} \setlength{\parskip}{0in}
\setlength{\topsep}{0.0in} \setlength{\partopsep}{0in}
\setlength{\leftmargin}{0.2in}}}{\end{list}}
\renewcommand{\familydefault}{\sfdefault}
%\usepackage[sfdefault, light, condensed]{roboto}
\usepackage[light,condensed]{iwona}
\usepackage[T1]{fontenc}
\usepackage[colorlinks=false,urlcolor=magenta,citecolor=blue,linkcolor=blue]{hyperref}
\usepackage{setspace}
\usepackage[dvipsnames]{xcolor}
\begin{document}
\begin{spacing}{0.98}
\name{{\huge\bf Jon Zink, Ph.D.}\vspace*{.1in}}
\begin{resume}
\vspace*{-2mm}
\section{Basic\\Information}
\begin{tabular}{@{}p{4.75in}p{4in}}
(248)909-1107 & \href{https://www.jonzink.com}{jonzink.com} \\
\href{mailto:[email protected]}{[email protected]} & \href{https://github.com/jonzink}{github.com/jonzink} \\
& \href{https://www.linkedin.com/in/jon-zink-phd/}{linkedin.com/in/jon-zink-phd} \\
\end{tabular}
% \vspace*{-0.2in}
\vspace{-3mm}
\textbf{\textit{\color{RoyalBlue} Computational astrophysicist with a proven history of using statistical analysis to characterize exoplanet systems looking for a Data Science position in a mid-sized data-centric company.}}
\vspace{-3mm}
% Collaborative, scientific thinker passionate about discovering and communicating nuanced insight from complicated data. Background includes: open-source contributions, project leadership, computer vision, traditional machine learning, and working with large, heterogeneous, often noisy datasets.
\section{Professional \newline Experience}
% \textbf{Activision Publishing, Inc.} \hfill Boulder, Colorado\newline
% \textit{Senior Machine Learning Engineer} \hfill \textbf{January, 2021 - Present}\newline
% As part of the Advanced Analytics and Machine Learning (ML) team, I support all ML initiatives for the Call of Duty and Warzone franchises. This includes being involved in early concept designs through the productionalization of mature models and stakeholder management. Below are selected highlights from this very diverse role:
% \begin{list2}
% % \vspace*{-5mm}
% \item Designed, built, maintained, and improved ML infrastructure. This includes tool developement like autoML, CI/CD (Jenkins), model tracking and management (MLflow) and orchestration (Airflow).
% \item Crafted and implimented policy around external data ingestion and data product retention for both my internal team and for sharing across the organization.
% \item Designed, built, and maintained near-realtime applications (via Spark streaming) to combat in-game cheating and other malicious behavior. Typical time-to-action is in the low 10s of seconds.
% \item Lead the transition of ML infrastructure from AWS to GCP/GCS.
% \item Created ML models to provide insights into customer conversion, churn, and behavioral segmentation. This leveraged survival analysis, clustering, as well as tree-based and linear methods.
% \item Supervised junior team members to design and develope recommendation systems to be productionalized in an upcoming title.
% \end{list2}
% \vspace*{-2mm}
%
% \textbf{Insight Data Science} \hfill New York, New York\newline
% \textit{Fellow} \hfill \textbf{January, 2020 - 2021}\newline
% \begin{list2}
% \vspace*{-5mm}
% \item Helped optimize the way NYC health inspectors perform restaurant inspections in order to reduce the time critical health violations remain unaddressed.
% \item Trained a random forest in Python to prioritize NYC restaurant inspections based on environmental variables and their past inspection histories and provided the results to NYC through an API deployed on AWS.
% \item Resulted in NYC inspectors identifying $\sim$2.5\% more violations in the first half of an inspection window, leading to critical violations being discovered up to 7 days earlier than by the current approach implemented by NYC.
% \end{list2}
% \vspace*{-2mm}
\textbf{California Institute of Technology} \hfill \newline
\textit{NASA Hubble Postdoctoral Fellow} \hfill \textbf{September, 2021 - Present}\newline
\begin{list2}
\vspace*{-5mm}
\item Led and coordinated team of 16 scientists, including junior researchers, to perform quality control tasks; to write grants; deliver science products; and produce peer-reviewed publications.
\item Extracted unprecedented exoplanet population trend with 99.99\% confidence by accounting for sample biases/reliability using Markov Chain Monte Carlo methods: see \href{https://exoplanets.nasa.gov/news/1768/discovery-alert-on-our-galaxys-outskirts-a-poverty-of-planets/}{NASA press release}.
% . I showed that planets are not uniformly distributed about the Milky-way and was highlighted in a \href{https://exoplanets.nasa.gov/news/1768/discovery-alert-on-our-galaxys-outskirts-a-poverty-of-planets/}{NASA press release}.
% \item Utilized statistical methods to extract population trends with 99.99\% confidence from an exoplanet dataset, revealing the non-uniform distribution of planets within the Milky Way galaxy, subsequently recognized by NASA and highlighted in a press release.
\item Executed K-means clustering analysis with Python/Scikit-learn to group planet classes, reducing the measurement uncertainty of existing methods by 22\%.
\item Implemented Bayesian hierarchical modeling with PyMC to resolve a 30-year-old planetary formation mystery by identifying an intra-system correlation in giant planet systems.
\item Developed Gaussian process regression methodology for robust correlated-noise removal, improving planetary signal recovery by 9X and adopted by 12 independent research groups.
\item Validated 60 planet signals through conditional survival analysis in R, concluding that each event had a false-positive likelihood of less than 0.1\%.
% Identified an intra-system correlation in giant planet systems using Bayesian hierarchical modeling. This work resolved a 30 year old mystery on the origin of Hot Jupiters.
\end{list2}
\vspace*{-0.08in}
\textbf{University of California, Los Angeles} \newline
\textit{Graduate Researcher} \hfill \textbf{September, 2016 - June, 2021}\newline
\begin{list2}
\vspace*{-5mm}
\item Developed Python algorithm able to search highly contaminated time series data and identify planet signals with 94\% reliability, resulting in the discovery of 372 planets: see \href{https://www.forbes.com/sites/jamiecartereurope/2021/06/11/we-found-372-new-alien-planets-using-a-long-dead-telescope-say-scientists/}{Forbes} \& \href{https://www.newsweek.com/astronomers-discover-366-new-worlds-gas-giants-kepler-k2-exoplanets-nasa-1653254}{Newsweek} press.
\item Implemented random forest regression with TensorFlow to characterize $\sim200,000$ stars, resulting in 93\% accuracy of the validation set classification, a 48\% improvement of existing methods.
% Executed memory intensive simulation 12X faster than previous researchers by applying parallel code structure to open-source code.
\item Derived Poisson point-process expansion, addressing bias issues associated with order statistics, improving the precision of inferred exoplanet population models by 36\%.
\item Produced novel likelihood function for forward model comparison, using the Anderson-Darling EDF and a modified Poisson PDF, enabling algorithm convergence 10X faster than existing methods.
% \item Awarded \$300k NASA grant for independent research.
\end{list2}
\vspace*{-2mm}
\section{Software Development}
\begin{list2}
\item Produced open source Python package able to parse exoplanet signals from statistical fluctuations in time series data, used by $27$ independent research groups: see \href{https://github.com/jonzink/EDI_Vetter_unplugged}{EDIunplugged} on Github.
\item Developed forward modeling software in Python and R (\href{https://github.com/jonzink/ExoMult}{ExoMult}), enabling direct assessment, via sample biased Monte Carlo simulations, of complex intra-system correlations.
% , reducing the uncertainty of previous Earth-analog occurrence calculations by 63\%.
\item Applied multi-core parallel structure to open-source code, resulting in a 12X speed boast for a memory intensive simulation.
\item Contributed to open source Python software through bug fixes and feature additions: see \href{https://github.com/California-Planet-Search/KPF-CPS}{KPF pipeline} on GitHub.
\end{list2}
\vspace*{-4mm}
% \textbf{Dept. of Physics and Astronomy, Texas A\&M University} \hfill College Station, Texas\newline
% \textit{Ph.D Candidate} \hfill \textbf{August, 2010 - 2016}\newline
% \begin{list2}
% \vspace*{-5mm}
% \item Demonstrated that measurements from a planned large observation campaign could be improved by up to a factor of 3 over traditional statistical methods through the use of machine learning.
% \item Implemented these machine learning methods and produced reliable results in a pilot survey of the real sky and under real-world conditions.
% % \item Collaborated with group members both in person, and through collaborative tools (e.g., GitHub, SVN).
% % \item Presented scientific results in high-impact, astrophysical journals and at international conferences.
% \end{list2}
% \vspace*{-2mm}
% \textbf{The University of Tennessee}, Knoxville, Tennessee USA\newline
% \textit{Master's Candidate} \hfill \textbf{August, 2007 - 2009}\newline
% \begin{list2}
% \vspace*{-5mm}
% \item Implemented a C-based pipeline to process hundreds of GBs of simulation results. Including a computer vision algorithm to automatically analyze and compare results to expected targets.
% \item Optimized simulation parameters using a genetic algorithm based search utilizing HPC (100k+ core) systems at the National Center for Computational Science, part of Oak Ridge National Laboratory
% \end{list2}
% \vspace*{-3mm}
% \section{Awesome Projects}
% \textbf{Using Imaging to Infer Galaxy Properties}\newline
% \begin{list2}
% \vspace*{-5mm}
% \item Predicted galaxy chemical composition with $\sim$5\% error from pseudo-three color imaging, a result better than other current, similar efforts in the literature. Leveraged CNNs to analyze $\sim$150,000 images of galaxies.
% \item Project start to publication: 4 months (typically $\sim$1.5 years). See: \href{https://github.com/boada/galaxy-cnns}{github.com/boada/galaxy-cnns}.
% \end{list2}
% \vspace*{-3mm}
% \textbf{Predicting Tournament Performance in Warmachine}\newline
% \begin{list2}
% \vspace*{-5mm}
% \item Created an \href{https://en.wikipedia.org/wiki/Elo_rating_system}{Elo} based model to forecast the results of upcoming tournaments and identify potential upsets.
% \item Integrated predictions into a local community ranking system and forecasted $\sim$1800 tournament game results of the popular tabletop game using Python (e.g., Pandas).
% \end{list2}
% \vspace*{-1mm}
\section{Skills}
\textbf{Machine Learning:} SVM, PCA, Supervised Learning, and Deep Learning\\
% \textbf{Statistical Methods:} Error Analysis and Hypothesis Testing\\
\textbf{Software and Computing:} SQL, HTML, multi-threading, and Object-Oriented Coding \\
\textbf{Professional:} 17 publications and 43 presentations for both technical/non-technical audiences
% \textbf{Leadership:} Experience organizing and leading workshops and collaboration meetings, Teaching and mentoring junior team members, Eagle Scout. \\
\vspace*{-3mm}
% \noindent\rule{8cm}{0.4pt}
% \fullline{ % hrules only listen to \hoffset
% \nointerlineskip % so I have this code
% \moveleft\hoffset\vbox{\hrule width\textwidth}
% \nointerlineskip
% }
%section for two column education
\section{Education}
\begin{tabular}{@{}p{3in}p{3in}}
\textbf{University of California}, Los Angeles
\begin{list2}
\item Ph.D., Astrophysics, 2021
\item B.S., Astrophysics, 2014
\end{list2} &
\end{tabular}
\vspace*{-4mm}
\end{spacing}
\end{resume}
\end{document}