Xiangying (Linda) Meng Email email@example.com
Location Troy, Michigan
A motivated data scientist with 10+ years of experience developing data-driven solutions for comprehensive assessment and analysis of high-volume data. Received multidisciplinary training in mathematics, applied mathematics, and dynamic engineering. Routinely worked on mathematical modeling and comprehensive data analysis such as advanced statistical analyses with cross validation, predictive data modeling, supervised and unsupervised learning to uncover complex results. Capable of executing large scale projects, supervising students and trainees and communicating with people from different backgrounds. SUMMARY
Specialized in mathematical modeling, data processing, data visualization and interpretation, statistical modeling.
Completed 26 federal/state projects in transforming high volumes of complex biological data into analytical solutions to uncover neuron network activity, generated reports and published 24 papers in high impact journals.
Created a new data processing platform for quantitative analysis of biological data, now used in over 10 groups.
Daily work with descriptive/predictive modeling and data wrangling tools (SQL, Python, MATLAB, R, SAS), proficient in feature engineering for structured and unstructured data.
Create weekly data-driven report to support team decision making, currently lead a team of 3 members.
Passionate about machine learning and AI, obtained several certificates and enjoyed daily learning. SKILLS
Python (Numpy, Pandas, Matplotlib, Tensorflow, Keras, Sklearn), MySQL, MATLAB, R, SAS, C Techniques Mathematical modeling, Regression, classification, clustering, time series analysis, predictive modeling, cross validation, supervised & unsupervised learning, image processing, statistics (Bayes’ theorem, Shannon entropy, Mutual information etc.) Tools Classification: logistic, k-means, decision trees, SVM, CNN, Spectral Clustering, Random forest, etc Object detection: YOLO, auto-encoder
Natural language processing: RNN, attention model, LSTM, vectorization WORK EXPERIENCE
Research Associate (06/2020-)
Neuro-engineering (supervisor: Dr. Patrick Kanold), Johns Hopkins University
Transform high volumes of unstructured complex biological data into computational models by applying information theory (maximum likelihood estimation, bayes theorem, entropy, optimization) and machine learning methods (regression fitting, dimension reduction, deep learning network).
Create data-driven solutions with data wrangling tools and descriptive/predictive modeling (Python, MATLAB, MySQL, SAS, etc.) to uncover neuron network activity. Post-doctoral Research Associate (09/2011-05/2020) Neuro-engineering (supervisor: Dr. Patrick Kanold), University of Maryland
Created a novel data processing platform for quantitative analysis of large datasets of unstructured biological- data (Github: https://github.com/lindamengxy/LSPSTOOLBOX.git). The GUI is able to effectively extract the features of the dataset and automatically generate ready-to-publish figures. It allows people with no programming background to be able to analyze complex data, visualize the data features and generate statistically meaningful results. It has become a popular data processing tool in computational neuroscience, used in over 10 research groups.
Built efficient image analysis workflows using dimension reduction methods such as PCA and auto encoder to provide rigorous, quantitative description of imaging time-series datasets.
Created multilayer neural networks to characterize large volumes of image and neuron activity data. The model can precisely and efficiently predict selectivity both qualitatively and quantitatively.
Developed predictive models for categorization of high volumes of neuron activity data: applying unsupervised hierarchical cluster analysis and deep learning to characterize data and separate them into multi distinct classes.
Designed auto calibration system to drive high-powered lasers with machine learning techniques.
Supervised 8 undergraduate students and 1 graduate student in applying data science tools. Research Assistant (09/2009 – 02/2011)
Mathematical Neuroscience (supervisor: John Rinzel), New York University
Developed mathematical models for simulating brain activities.
Designed neural network for neural encoding.
Applied advanced mathematical analysis methods to better understand underlying biologicalphysical mechanisms.
2011 Beihang University, PhD, Engineer (Mathematical Neuroscience)
(co-supervised by Dr. John Rinzel at New York University & Dr. Qishao Lu at Beihang University) 2007 Hebei Normal University, MA, Mathematics
2004 Hebei Normal University, BS, Mathematics and Applied Mathematics CERTIFICATE
Machine learning by Stanford University on Coursera
Convolutional neural networks, Coursera
Structuring machine learning projects, Coursera
Neural networks and deep learning, Coursera
Improving deep neural networks: hyperparameter tuning, regularization and optimization
R programming, Johns Hopkins University, Coursera
Chinese National Computer Rank Examination 2 and 3 for C AWARDS
INC Travel Award, International Neural Coding 2009, Taiwan, May 2009
Accepted as one of 20 outstanding international students (first student from mainland China) to work with world well-known professors to build up intellectual neural networks, Advanced Course in Computational Neuroscience, Freiburg, Germany, August, 2010
AIM Travel Award. Being invited as one of 28 mathematicians to work in group with famous professors to solve interesting problems in neural networks with mathematical tools, Stochastic Dynamics of Small Networks of Neurons, Palo Alto, California, February, 2012
MBI Travel Award, Sensory Systems and Coding, Columbus, Ohio, May, 2013
ARO Travel Award, one of the 40 outstanding graduate students and postdocs selected based upon their recognized research achievements among ~2500 applications. 37th ARO Midwinter Meeting, San Diego, California, USA, 2014
Won the second scholarship for 4 consecutive years during college (2000-2004) ADDITIONAL ACTIVITIES
Built CNN ResNet for face reorganization with Keras.
Applied pre-trained YOLO net for car detection.
Using pre-trained VGG network for neural style transfer.
Applied LSTM on music pattern generation, trigger word detection and Emoji generation from word text.
Attended Johns Hopkins’ TRIPODS Winter School & Workshop for deep learning. Hands-on experience on machine learning, graphs, and optimization (Jan 6th to 15th, 2021).