LZ Investments is an innovative real estate investment company that helps busy professionals like you build passive income. Our revolutionary investment concepts help you reduce risk and increase yield. At LZ Investments, our mission is to make wealth building through real estate investing possible for everyone, regardless of your risk tolerance.

RioConsulting is a clinical research organization (CRO) and a leading provider of clinical data science staffing solutions. With a dedication to bring innovative and customized data solutions, RioConsulting applies firm commitment to quality to our clients and partners.

1. Participated Freddie Mac Automatic Collateral Evaluator (ACE) project, which automatically evaluates loan quality and grants appraisal waiver. Developed the ACE prototype by Hadoop, HIVE and PySpark, visualized result by Plotly and R.2. Developed appraisal quality model in Loan Collateral Advisor (LCA) by logistic regression and feature selection.

1. Participated Freddie Mac Automatic Data Valuation System (ADVS) project. Major roles were rules engine establishment and rule writing (R code). ADVS implementation will significantly impact home mortgage industry by accelerating appraisal approval process and eliminating most of human review effort.2. Developed Cascade Valuation Model (CVM) to leverage Clear Capital internal valuation models and provide most accurate home valuations. 3. Improved Clear Capital Comp Ranking algorithm by employing Principle Component Analysis (PCA) and automatic weights update algorithm. Designed survey tool to estimate quality of new algorithm by R shiny application.4. Independently designed and implemented Complexity Score Algorithm to identify potential complex properties and facilitate Clear Capital solicitation process.5. Systematically and repeatedly retrieved nationwide sales data from Multiple Listing Service (MLS) by implementing K means algorithm.

1. Data cleaning, transformation and visualization in R and Tableau. 2. Explored and identified underline driving reasons for Wall Street analysts to be successful (stardom), switch employees, get promotion and quit industry by logistic regression and conditional classification tree.3. Performed variance decomposition by modeling random effect of mixed model, and parameters were simulated by Markov Chain Monte Carlo (MCMC) algorithm.

1. Data cleaning, transforming and descriptive analysis in R.2. Analyzed longitudinal sales data by using linear mixed model and generalized linear model (poisson regression & negative binomial regression) in SAS. 3. Implemented sequential variance decomposition method to explain relative importance of sales-driven factors.4. Successfully validated analysis datasets, selected statistical model and identified critical sales-driven factors.

1. Followed current commodity market and global economics. Identified critical economic factors related to commodity market, such as GDP, employment rate. 2. Screened and Developed strategic commodity baskets based time-lag value correlations. 3. Designed and explored prediction models within each basket, including regularized linear regression, regression tree, PCA and random forest etc.4. Simulated profit return by leveraging long/short signals from commodity basket combinations.

1. Implemented expectation maximization(EM) algorithm and SAS multiple imputation for missing data. 2. Promoted monthly online sales predicting accuracy by random forest and gradient boosting algorithms.3. Successfully classified top-sale products using naive bayes, decision tree and support vector machine algorithms, and achieved high predicting efficiency.

1. Analyzed time series trend and periodic cycle for individual physical activity using kernel smoothing. 2. Developed R functions to extract activity-specific features from time series data. For example, extract dominate frequency by fast fourier transform(FFT). 3. Designed classification tree to predict the activity types, and regression tree to predict the energy expenditure. 4. Improved the predicting accuracy by using random forest algorithm. And achieved high predicting accuracy compared to other algorithms: Naive Bayes, Support Vector, and Neutral Network.