About Me.

Hello, my name is Keang Cheang Ung, a doctoral student at Colorado Technical University (CTU). I have a background in data science, economics, and the stock market. I graduated from Texas A&M University in 2017 and moved to Denver to continue my doctoral program in computer science. I also have a strong passion in the equity market and day-trading, and I started day-trading at home using my personal account in April.

The purpose of creating this blog is to 1) provide news and recent developments in artificial intelligence, and 2) to share knowledge and lessons related to AI, deep learning and data science in general. Also, another reason to start this blog is to fulfill an individual project in my class, the Futuring and Innovation at CTU. This class opens the door to new innovative ideas that can affect organizations and how the new trends in technology like AI and big data can change the future.

The theme in the blog will cover theories and applications of AI and data science. As a doctoral student in computer science focusing on big data, I have a research passion for building a predictive algorithm to forecast the direction of the stock price given the duration of the investment and risk involved. In order to conduct this research, I have to study prior literature that has been carried out with similar objectives of predicting the stock market movement. For example, Lachiheb and Gouider (2018) used four years of price data to forecast the future 5-minute price. The strength of this study was the fact that the researcher used the deep neural network (DNN) with dimension reduction technique, variable selections, and correlation factors incorporating into the model making it outperforms the traditional neural network like artificial neural network (ANN). Also, the price input data was long history that allows the models to learn and train to adapt patterns better compared to models using a short period of price data (Lachiheb, & Gouider, 2018).

Another interesting approach to forecasting the stock price direction is to use pattern recognition through neural networks and point reduction techniques. Chen and Chen (2016) built an intelligence model for analyzing and predicting the stock market using the perceptual important point (PIP) to reduce the unnecessary points in the time series data and then feed them to neural networks. The intelligence parts of this algorithm were to capture the bull-flag pattern in the technical chart with the least running time as possible and to predict the direction of the stock price with the high accuracy. The results showed that the intelligent algorithm proposed in this study shows a better return than the rough set theory (RST), genetic algorithms (GAs), and their hybrids. By correctly predicting the bull flag turning point, the investors could benefit huge returns by executing the stock orders in a timely manner.

In conclusion, I hope this blog will give values to you as someone is interested in AI and big data. We will apply machine learning and deep learning together to solve real-world issues, especially in the investment arena.

Reference:

Chen, T. L. & Chen, F. Y. (2016). An intelligent pattern recognition model for supporting investment decisions in stock market. Information Sciences, 346–347, 261-274. doi: 10.1016/j.ins.2016.01.079.

Lachiheb, O., & Gouider, M. S. (2018). A hierarchical deep neural network design for stock returns prediction. Procedia Computer Science, 126, 264-272. doi: 10.1016/j.procs.2018.07.260.