UROP Proceedings 2020-21

School of Engineering Department of Industrial Engineering and Decision Analytics 166 Department of Industrial Engineering and Decision Analytics Big Data Analytics Techniques for Quality and Process Improvement Supervisor: TSUNG Fugee / IEDA Student: MU Yiteng / DA Course: UROP1100, Spring Traffic flow forecasting has been more and more critical nowadays. In this report, we give a literature review on current research around this topic. Based on what we have learned from the past experience, we aim to build a model based on the historical data fromMTR and predict the traffic flow during the operation hours. We extract useful information from the raw data and use ARIMA (Autoregressive Integrated Moving Average) technique, which is a widely used model in this field. Moreover, we also tried different models for different types of stations to get the best performance of prediction. Finally, we did an evaluation of the current limitation and defects based on the forecasting results as well as the research progress. Data analysis in FinTech Supervisor: ZHANG Jiheng / IEDA Student: LIN Ruihan / DA WONG Siu Tim / DA Course: UROP1100, Summer UROP1100, Summer In this project, we aimed to develop an auto-trading strategy on cryptocurrency market using technical indicator combined with machine learning algorithms. We discovered two main technical indicator -- SAR & MACD to help catching the market opportunities, and used random forest classifier to reinforce the signal accuracy. By connecting to the server on IEDA department, we performed rigorous backtest and develop interactive plots for visualization. Lastly, we connect to the exchange by using external API -- CCXT to achieve real-time auto trading. Data analysis in FinTech Supervisor: ZHANG Jiheng / IEDA Student: XIAO Hanyu / SSCI Course: UROP1100, Summer In this report, we present some data analysis skills with python. We will first show how to use some python packages and library tools, such as pandas, regular expressions, scipy.stats, to process rough data and the process of getting CSV, HTML files, manipulate these data using regex (regular expressions). Next, through some useful data analytic tools in python, like aggregation, merge, group by, form a helpful data frame, and finally use scipy.stats tools to calculate the correlation index. Finally, we will also point out some flaws and limitations of this data visualization and put forward some proposed methods to improve it.