JOHN SHING CHAU LEUNG

+1-401-***-**** adii25@r.postjobfree.com

*** **** ******, **********, ** 02906

EDUCATION

Brown University Aug 2014 { May 2020

PhD in Physics, GPA: 4.0

Thesis: \Dark Matter Searches in the Crescendo of the 21st Century Experiments" Adviser: JiJi Fan Chinese University of Hong Kong Aug 2012 { Jun 2014 Master of Philosophy in Physics, GPA: 3.6

Thesis: \Topics in Cosmic Ray Measurement and Particle Detectors Development" Adviser: Ming Chung Chu Chinese University of Hong Kong Aug 2009 { Jun 2012 BSc major in Physics, minor in French, GPA: 3.6

Thesis: \Rabi Oscillation and State Construction in Quantum Bouncing Ball" Adviser: Chi Kwong Law SKILLS

Coding Python, C++, C, FORTRAN, Bash, SQL, VHDL, Verilog Softwares Mathematica, MATLAB, Tableau, ROOT, Keras, TensorFlow, Jekyll, LA TEX

OS Windows, Linux, MacOS

Languages English ( uent), Mandarin ( uent), Cantonese (native), French (intermediate), Italian (beginner) EXPERIENCE

Coursera Deep Learning and Machine Learning, independent projects based on coursework May 2020

(o ered by deeplearning.ai and Stanford University) CNN on hand gesture recognition using TensorFlow (github.com/bbsonjohn/tensor ow CNN)

{ Wrote a deep network that parses gestures to numbers using RGB-colored images with 12288 channels.

{ Trained on 1000 images result in 80% accuracy on test set. CNN on smile detector using Keras (github.com/bbsonjohn/keras CNN)

{ Created a CNN that reads in RGB-colored photos with 12288 channels of people’s faces.

{ Trained on 600 samples on predicting whether a person is smiling, resulting in 95% test set accuracy. RNN with LSTMs on text translation and synthesis using Keras (github.com/bbsonjohn/RNN keras)

{ Established RNNs based on LSTMs to translate French sentences to English.

{ RNN trained on 8000 sentences with up to 30 words, to translate French congressional texts to English.

{ Trained RNN samples sequences to synthesize new English sentences (tested up to output 60 words). Brown University, particle phenomenology research Aug 2014 { May 2020 Locating dark matter in the Milky Way

{ Investigated into a SQL database containing 100 million samples of stars’ kinematic data.

{ Used statistical and machine learning methods such as Bayesian statistics, na ve Bayes models, kernel density estimation and Markov-chain Monte-Carlo methods, written in Python, to constrain dark matter distribution in the Milky Way (Publication 2). Predictive study on the discovery potential of future high energy particle colliders

{ Monte-Carlo simulated and analyzed 1 billion samples of particle data of a size of 10 TBs using C++.

{ Utilizing hypothesis testing, forecast colliders ability to discover new physics models (Publication 3). Understanding astrophysics-induced uncertainties in dark matter detection

{ Analyzed the time-series signals of underground dark matter detector to discover new correlations between dark matter signals and stars’ movements (Publication 1). Lawrence Berkeley National Laboratory, detector front-end development Aug 2012 { May 2014 Developed VHDL and Verilog data acquisition rmwares on FPGAs to process detector data at a time-scale of 10 nanoseconds.

Chinese University of Hong Kong, detector development and data analysis Aug 2010 { May 2014 Constructed a muon telescope with a hardware control written in C. Coded a back-end C++ program for data cleaning and signal detection on time series data of the muon telescope’s pixels.

European Organization for Nuclear Research, detector development May 2011 { Aug 2011 Was the rst Hong Kong Universities representative to participate in research activities at CERN. SELECTED PUBLICATIONS

1.) J Buch, J Fan, and JSC Leung, \Implications of the Gaia sausage for dark matter nuclear interactions", Phys. Rev. D 101 (2020).

2.) J Buch, JSC Leung, and J Fan, \Using Gaia DR2 to constrain local dark matter density and thin dark disk", JCAP 1904 (2019).

3.) J Fan, P Jaiswal, and JSC Leung, \Jet observables and stops at a 100 TeV collider", Phys. Rev. D 96 (2017).