Renee S. Liu

Renee S. Liu

United States
1K followers 500 connections

Join to view profile

About

When you do things from your soul, you feel a river moving in you, a joy -- Rumi

I…

Contributions

Activity

Join now to see all activity

Experience

Education

  • University of California, Davis Graphic

    University of California, Davis

    -

    Ph.D. seeking; Specializing in two-sided markets and microeconometrics.

    Selected Ph.D. Coursework:
    Computational Statistics (PhD level Statistics Department) - Implement a set of computational techniques, such as randomized matrix multiplication, deterministic and stochastic optimization, for statistical analysis of big-data using R programming language
    Econometrics Foundation - Focus: Theoretical/statistical foundation for inference
    Econometrics Methods I - Focus: Applied…

    Ph.D. seeking; Specializing in two-sided markets and microeconometrics.

    Selected Ph.D. Coursework:
    Computational Statistics (PhD level Statistics Department) - Implement a set of computational techniques, such as randomized matrix multiplication, deterministic and stochastic optimization, for statistical analysis of big-data using R programming language
    Econometrics Foundation - Focus: Theoretical/statistical foundation for inference
    Econometrics Methods I - Focus: Applied cross-sectional and panel data methods
    Econometrics Methods II - Focus: Applied inference methods
    Cross Section Econometrics - Focus: Theoretical foundation for cross-sectional data and inference
    Cross Section Topics - Focus: Machine Learning, inference, large datasets
    Industrial Organization I - Focus: Theoretical foundation on Firm entry and
    competition
    Industrial Organization II & III - Focus: Empirical analysis of product differentiation
    and innovation

  • -

    Activities and Societies: A full year of hard work!

    12-month, complete mastery-based projects based on problems encountered in business application development. Full-time Java backend software engineering program co-developed by Kenzie Academy and Amazon Technical Academy.

    I developed adaptable skills to research, build, and deploy cloud-enabled software applications, while navigating the Software Development Lifecycle (SDLC) and collaborating in Agile Scrum teams.

    OOP, unit / integration testing, DynamoDB, and all the other fun…

    12-month, complete mastery-based projects based on problems encountered in business application development. Full-time Java backend software engineering program co-developed by Kenzie Academy and Amazon Technical Academy.

    I developed adaptable skills to research, build, and deploy cloud-enabled software applications, while navigating the Software Development Lifecycle (SDLC) and collaborating in Agile Scrum teams.

    OOP, unit / integration testing, DynamoDB, and all the other fun things. Developed a capstone project with Sprintboot framework and AWS Lambdas & DynamoDB, allowing users to interact with the backend through open APIs.

  • -

    Train and deploy models using FastAPI and Heroku with CICD process.

  • -

    1. (Self directed Capstone Project) I built a command-line stock price predictor using LSTM model and argparse package. The end product allows a user to input the name of the stock she/he is interested in from the terminal, and this application will predict n-days of prices to help this user make buy-sell decisions.
    2. Implemented feature engineering techniques -- containment and Longest common sequence -- using a corpus of texts data to classify if each document is a result of plagiarism…

    1. (Self directed Capstone Project) I built a command-line stock price predictor using LSTM model and argparse package. The end product allows a user to input the name of the stock she/he is interested in from the terminal, and this application will predict n-days of prices to help this user make buy-sell decisions.
    2. Implemented feature engineering techniques -- containment and Longest common sequence -- using a corpus of texts data to classify if each document is a result of plagiarism or not. I trained a custom sklearn logistic regression model and deployed the trained model onto AWS using SageMaker. The accuracy is around 95% but it is due to small training dataset.
    3. Deployed a sentiment analysis model on AWS SageMaker with Lambda and API Gateway. This final product is a web application that uses the deployed model to predict a user-supplied movie review.

  • -

    1. Built relational data models with Postgres
    2. Modeled event data to create a NoSQL database and ETL pipeline for a music streaming
    app. They will define queries and tables for a database built using Apache Cassandra.
    3. built an ETL pipeline that extracted data from AWS S3, staged them in AWS Redshift, and transformed data into a set of dimensional tables for downstream data consumers.
    4. built a data lake and an ETL pipeline in Spark that loads data from S3, processes the data…

    1. Built relational data models with Postgres
    2. Modeled event data to create a NoSQL database and ETL pipeline for a music streaming
    app. They will define queries and tables for a database built using Apache Cassandra.
    3. built an ETL pipeline that extracted data from AWS S3, staged them in AWS Redshift, and transformed data into a set of dimensional tables for downstream data consumers.
    4. built a data lake and an ETL pipeline in Spark that loads data from S3, processes the data into
    analytics tables, and loads them back into S3.
    5. Scheduled a complete data ETL pipeline by creating and automating a set of data pipelines with Airflow, monitoring and debugging production pipelines.

  • -

    Used Pytorch to implement CNN and RNN models;
    Deployed an RNN model backed web application to predict movie review sentiments using AWS Sagemaker.

  • -

    Successfully built machine learning models including supervised and unsupervised methods; created and run data pipelines; built recommendation systems; deploy deep learning models to web applications.
    Completed 6 data science projects with real-world dataset.

  • -

    Streaming data to and from Kafka broker using Pyspark.

  • -

  • -

  • Completed 5 projects that solve real world problems using CNN, RNN, and various deep learning techniques.

  • -

    1. Studied the fundamentals of distributed computing with Spark on DataBricks platform.
    2. Trained a logistic regression model with Spark SQL on a distributed large dataset.
    3. Manipulated distributed dataset using Spark SQL.

Licenses & Certifications

Projects

  • Command Line Stock Price Predictor with LSTM model

    - Present

    Built an easy-to-use command-line stock predictor that can make predictions on a particular stock's prices on multiple days in the future. The underlying model is a 2-layer LSTM neural network. It is simple yet very effective in comparison to other more traditional time series forcasting models.

    Once you have fired up your terminal, the only thing you need to type in is the ticker name of the stock of your interest. Optionally, you can type in how many days' predictions you want to see…

    Built an easy-to-use command-line stock predictor that can make predictions on a particular stock's prices on multiple days in the future. The underlying model is a 2-layer LSTM neural network. It is simple yet very effective in comparison to other more traditional time series forcasting models.

    Once you have fired up your terminal, the only thing you need to type in is the ticker name of the stock of your interest. Optionally, you can type in how many days' predictions you want to see in the future. Finally, you will see a list of price predictions of this stock. The goal is to quickly help you make buy&sell decisions. (Note, I will not be responsible to any monetary loss if you happen to use this application!).

    See project

  • My coding gists for SQL, Python, and R

    - Present

    See project

  • Implementing Expectation-Maximization Algorithm in R

    See project

  • Design OLAP data model to build a cloud-based data warehouse in Snowflake

    -

    Ingest and migrate YELP business data into Staging, Operational Data Store (ODS), and Data Warehouse environments in Snowflake. Then ultimately query the data for relationships between weather and Yelp reviews.

  • Dog Breeds Classification using Deep Learning with Pytorch

    -

    I used transfer learning technique with Pytorch library to employ the well-trained CNN feature extraction layers from VGG16, then I built and trained a self-defined fully-connected linear classifier layer to complete an image classification neutral network.

    I reached 80% accuracy on the first try.

  • A Web Application powered by CNN InceptionV3 model to Classify Dog Images

    -

    Using 8651 Dog Images to train a Convolutional Neural Network through a transfer learning technique, I am able to reach an accuracy of 82% when predicting a new dog image. I then deployed this image classifier as a Web App using Flask microframework. The web application is able to take in a user-uploaded dog image and predict the breed of this dog within 5 min. Not so fast but it works well.

    See project

  • Article Recommendation Engine

    -

    Building a collaborative-filtering based recommendation engine.
    The final goal of this project is to deploy this engine to a web application.

  • Create Customer Segmentations

    -

    In this project, I applied unsupervised learning techniques on demographic and spending data for a sample of German households data provided by Arvoto Financial Services to identify customer segments hidden in the data.
    I preprocessed the messy data, applied dimensionality reduction techniques - PCA, and implemented k-mean clustering algorithms to segment customers with the goal of identifying the best segment of customers to outreach for a mail-order company to minimize the mailing…

    In this project, I applied unsupervised learning techniques on demographic and spending data for a sample of German households data provided by Arvoto Financial Services to identify customer segments hidden in the data.
    I preprocessed the messy data, applied dimensionality reduction techniques - PCA, and implemented k-mean clustering algorithms to segment customers with the goal of identifying the best segment of customers to outreach for a mail-order company to minimize the mailing costs.
    Finally, the result allowed me to compare the segmentation found with additional labeling and to consider ways this information could assist the wholesale distributor with future service changes.

  • Deploy a Web Application to Linux Server using AWS Lightsail

    -

    Built and deployed a Web Application to Linux Server. Secured the remote server and database server.

    See project

  • Image Classifier

    -

    Implemented an image classification application using a deep learning model via Pytorch on a
    dataset of images. Utilized transfer learning model - VGG16 - to build a neural network that can classify a user-input flower image into top 5 most likely flower categories. Finally, I converted the trained model into a Python application (with python Argparse package) that anybody can run from the command line.

Recommendations received

1 person has recommended Renee S.

Join now to view

More activity by Renee S.

View Renee S.’s full profile

  • See who you know in common
  • Get introduced
  • Contact Renee S. directly
Join to view full profile

Other similar profiles

Explore more posts

Explore collaborative articles

We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.

Explore More

Add new skills with these courses

See all courses