George Liu

Using state of the art GAN's to generate human face images based on the CelebFaces Attributes Dataset (CelebA) dataset.

Training a sequence to sequence model on a dataset of English and French sentences that can translate new sentences from English to French

In this project, we generate Simpsons TV scripts using RNNs. Part of the Simpsons dataset of scripts from 27 seasons are used. The Neural Network generates a new TV script for a scene at Moe's Tavern.

In this project, we classify images from the CIFAR-10 dataset. The dataset consists of airplanes, dogs, cats, and other objects. After preprocessing the images, we train a convolutional neural network on all the samples.

In this project, we build a neural network with Numpy and use it to predict daily bike rental ridership.

Quora uses a combination of machine learning algorithms and moderation to ensure high-quality content on the site. High answer quality has helped Quora distinguish itself from other Q&A sites on the web. In this project, we will devise a classifier that is able to tell good answers from bad answers, as well as humans can.

--> Technology Used: Python, Scikit Learn, xgboost.
From frontline support teams to C-suites, customer satisfaction is a key measure of success. Unhappy customers don't stick around. What's more, unhappy customers rarely voice their dissatisfaction before leaving. In a recent Kaggle competition, Santander Bank asks Kagglers to help them identify dissatisfied customers early in their relationship. Doing so would allow Santander to take proactive steps to improve a customer's happiness… Show more

--> Technology Used: Python, Scikit Learn, xgboost.
From frontline support teams to C-suites, customer satisfaction is a key measure of success. Unhappy customers don't stick around. What's more, unhappy customers rarely voice their dissatisfaction before leaving. In a recent Kaggle competition, Santander Bank asks Kagglers to help them identify dissatisfied customers early in their relationship. Doing so would allow Santander to take proactive steps to improve a customer's happiness before it's too late. In this project, we'll tackle this competition and try to predict whether a customer is satisfied or not using customer data. Show less

--> Technology Used: Python, Scikit Learn.
In this project, we aim to use unsupervised learning techniques to help our client, a wholesale grocery distributor, to identify different customer segments in order to better understand customer behaviors and devise corresponding marketing strategy and operational plan to better meet customers’ needs. Principal Component Analysis (PCA), Independent Component Analysis (ICA) and Gaussian Mixture Model clustering (GMM) techniques are explored and… Show more

--> Technology Used: Python, Scikit Learn.
In this project, we aim to use unsupervised learning techniques to help our client, a wholesale grocery distributor, to identify different customer segments in order to better understand customer behaviors and devise corresponding marketing strategy and operational plan to better meet customers’ needs. Principal Component Analysis (PCA), Independent Component Analysis (ICA) and Gaussian Mixture Model clustering (GMM) techniques are explored and final customer segmentations are recommended. Show less

--> Technology Used: Python, Scikit Learn, Numpy.
A smart cab is a self-driving car from the not-so-distant future that ferries people from one arbitrary location to another. In this project, we will use reinforcement learning to train a smart cab drive. In particular, the Q-learning algorithm will be implemented in a simulated environment to learn the optimal actions to take for the cab to drive to a certain target location.

The simulated environment is a grid world comprising a… Show more

--> Technology Used: Python, Scikit Learn, Numpy.
A smart cab is a self-driving car from the not-so-distant future that ferries people from one arbitrary location to another. In this project, we will use reinforcement learning to train a smart cab drive. In particular, the Q-learning algorithm will be implemented in a simulated environment to learn the optimal actions to take for the cab to drive to a certain target location.

The simulated environment is a grid world comprising a primary driving agent (the smart cab) and several dummy agents as traffic. There are traffic lights at all intersections, the primary driving agent needs to learn to drive according to traffic and light situations. Show less

--> Technology Used: Python, Scikit Learn.
With the advent of the data analytics era, student learning data is becoming increasingly
available through various online education tools such as Edmodo etc. The availability of these
data, when coupled with powerful machine learning techniques, will provide invaluable insights
that were never possible before. The aim of this project is to build a student intervention
system that uses high school student data to make predictions… Show more

--> Technology Used: Python, Scikit Learn.
With the advent of the data analytics era, student learning data is becoming increasingly
available through various online education tools such as Edmodo etc. The availability of these
data, when coupled with powerful machine learning techniques, will provide invaluable insights
that were never possible before. The aim of this project is to build a student intervention
system that uses high school student data to make predictions about their likelihood of passing
final exam, so that appropriate early intervention can be made to ensure student success. Show less

--> Technology Used: Python, Scikit Learn.
Using historical data to predict Boston house prices.

--> Technology Used: Python, MongoDB.
OpenStreetMap(OSM) is an open map database built by volunteers around the world. The data created are available for free use under the Open Database License. As the data were crowd-sourced, there can be inconsistency in how the data is presented. In this project, we will improve the data quality with data wrangling and store the cleaned data in MongoDB. In particular, XML map data will be parsed, audited and processed for street type inconsistency… Show more

--> Technology Used: Python, MongoDB.
OpenStreetMap(OSM) is an open map database built by volunteers around the world. The data created are available for free use under the Open Database License. As the data were crowd-sourced, there can be inconsistency in how the data is presented. In this project, we will improve the data quality with data wrangling and store the cleaned data in MongoDB. In particular, XML map data will be parsed, audited and processed for street type inconsistency before being reshaped and converted into JSON format data, later imported into the MongoDB database. Show less

--> Technology Used: Python, Pandas, Numpy.
RMS Titanic's sinking was one of the worst maritime disasters in modern history. With the dataset obtained from Kaggle, we can now garner some insights about the passengers on board the ship. In this project, we will examine the Titanic dataset and try to answer the following questions: Were all passengers on board equally likely to survive? If not, what were some characteristics for people who survived compared with people who didn't?

--> Technology Used: R, dplyr.
In this project, we will research the 2012 General Social Survey data. Specifically, we’ll study the relationship between the “degree” and the “abnomore” variables to answer the question: “Is there a relationship between people’s education and whether they support abortions for women who are married and don’t want any more children?”

The General Social Survey (GSS) is a national research project intended to identify attributes and attitudes of the… Show more

--> Technology Used: R, dplyr.
In this project, we will research the 2012 General Social Survey data. Specifically, we’ll study the relationship between the “degree” and the “abnomore” variables to answer the question: “Is there a relationship between people’s education and whether they support abortions for women who are married and don’t want any more children?”

The General Social Survey (GSS) is a national research project intended to identify attributes and attitudes of the American society and to facilitate comparison between the US and other countries. The research scope includes respondents’ background, personal and family information, societal concerns, workplace and economic concerns etc. The survey covered the period from 1972 to 2012 (in this data set). Show less

In a Stroop task, participants are presented with a list of words, with each word displayed in a color of ink. The participant’s task is to say out loud the color of the ink in which the word is printed. The task has two conditions: a congruent words condition, and an incongruent words condition. In the congruent words condition, the words being displayed are color words whose names match the colors in which they are printed: for example RED, BLUE. In the incongruent words condition, the words… Show more

In a Stroop task, participants are presented with a list of words, with each word displayed in a color of ink. The participant’s task is to say out loud the color of the ink in which the word is printed. The task has two conditions: a congruent words condition, and an incongruent words condition. In the congruent words condition, the words being displayed are color words whose names match the colors in which they are printed: for example RED, BLUE. In the incongruent words condition, the words displayed are color words whose names do not match the colors in which they are printed: for example PURPLE, ORANGE. In each project, we measure the time it takes to name the ink colors in equally-sized lists. Each participant will go through and record a time from each condition. Show less

--> Technology Used: R, dplyr, caret.
Using devices such as Misfit and Fitbit it is now possible to collect a large amount of data about personal activity relatively inexpensively. One thing that people regularly do is to quantify how much of a particular activity they do, but they rarely quantify how well they do it. In this project, we try to identify human activity using data from accelerometers on the belt, forearm, arm, and dumbell of 6 participants.

--> Technology Used: R, Dplyr.
In this project, we look at the national storm data produced by the National Weather Service. The data covered period from 1950 to 2011. In particular, the data are used to identify the events that are most harmful to population health and the events that have the greatest economic consequences.

--> Technology Used: HTML, CSS, Javascript, D3.js, Dimple.js.
This visualization provides a quick overview of United States air carriers' flight delay situation. It highlights the percentages of delays and average delay time across different carriers. The visualization is based on RITA's 2008 flight delays data (the data is first transformed and summarized using R). UniqueCarrier (carrier name) and ArrDelay (arrival delay time in minutes) are the two variables that are examined for… Show more

--> Technology Used: HTML, CSS, Javascript, D3.js, Dimple.js.
This visualization provides a quick overview of United States air carriers' flight delay situation. It highlights the percentages of delays and average delay time across different carriers. The visualization is based on RITA's 2008 flight delays data (the data is first transformed and summarized using R). UniqueCarrier (carrier name) and ArrDelay (arrival delay time in minutes) are the two variables that are examined for insights in this project. Show less

A/B test is an effective and powerful tool in web site optimization and app development. It is widely used by data scientists in the technology industry. In this project, we will go over the full process of A/B testing and design an A/B test for Udacity in order to make a decision about the launch of a Free Trial Screener.

--> Technology Used: Python, Scikit Learn.
Once a gigantic corporation claiming to have revenues of 111 billion dollars, Enron’s share price decreased from 90 dollars to just pennies in mid-2000 before it filed for bankruptcy in 2001. The Enron Scandal was a result of “creative accounting” and corporate fraud, and this fraud was linked to a small number of employees out of Enron’s 20,000 staff. Given data, is it possible to detect fraud and identify “person of interest” for… Show more

--> Technology Used: Python, Scikit Learn.
Once a gigantic corporation claiming to have revenues of 111 billion dollars, Enron’s share price decreased from 90 dollars to just pennies in mid-2000 before it filed for bankruptcy in 2001. The Enron Scandal was a result of “creative accounting” and corporate fraud, and this fraud was linked to a small number of employees out of Enron’s 20,000 staff. Given data, is it possible to detect fraud and identify “person of interest” for investigation? In this project, we will use the publicly available Enron dataset that contains financial and email data of 146 people to identify fraud using machine learning techniques. Show less

--> Technology Used: R, ggplot2.
We have always relied on wine experts who use their esoteric jargon to rate wine qualities for us. But what exactly is wine quality based on? What are the criteria? In this project, we look at the Wine Quality dataset and use data analysis methods with R to explore the relationship between wine quality and various attributes such as acidity, sugar and alcohol.

--> Technology Used: R, tm, dplyr.
Yelp.com is available in 15 languages and has 142 million unique visitors per month. As a result, a huge amount of customer review data is generated on a regular basis. How can we turn the huge amount data generated by Yelp into business insights that are beneficial for the business community? In this project, we try to answer this question by examining data provided by the Yelp Data Set Challenge to garner business insights.

The project aimed at identifying the optimal growth strategy for Lansing Dental Center.

The project aimed at identifying improvement opportunities of the MCIS call center operations.