Finding the needle in a haystack of data
Professor Bijan Raahemi from the Telfer School of Management at the University of Ottawa has been awarded a Discovery Grant from the Natural Sciences and Engineering Research Council (NSERC) to design, verify and analyze algorithms in the areas of healthcare, corporate finance, and engineering.
Professor Raahemi will work with high dimensional data, where the data is described using a wide range of attributes. To put it into perspective, high dimensional data is all of the information that Facebook, for example, gather about its users; from their demographics, gender to their interests, travels, and interactions with other users.
Professor Raahemi will create and test algorithms to discover anomalies in this high dimensional data. ‘‘When we look for anomaly, we search for something that is not statistically the same as the majority of data, meaning that it does not fit “normal” patterns. If you see a data point that exhibits different behaviour than the majority of data, you found an anomaly’’ specifies professor Raahemi. Detecting anomaly in high-dimensional data especially in the presence of noise is an important problem and it represents a challenge in real-life situations.
In the field of healthcare, population health will be the focus of professor Raahemi’s program where he will try to monitor what is going on with a disease spreading. For example ‘‘with big data we can monitor the characteristic of flu season, how it is spreading, for how long and, where it is going’’ explains professor Raahemi. Another interesting application is to analyze vital signals collected from patients including temperature, heartbeat, blood pressure, and ECG signals for anomaly detection. The data will be collected from various sensors in real time and stored in a cloud.
In the corporate finance area, professor Raahemi will look for the anomalies manifesting as fraudulent financial transactions. Fraudulent activities have been a major concern for banks as they impact the bottom line and also are inconvenient for bank’s customers when their credit card or debit card are “hacked” and used for unauthorized purchases.
The third area that professor Raahemi will explore is engineering, more specifically the Internet traffic. The amount of traffic on the Internet on a daily basis is incredibly high and hackers can cripple businesses and organizations by attacking their websites with so much traffic that it becomes unresponsive and shuts down. Professor Raahemi wants to create an algorithm that will help shield business and organizations from these attacks.
This innovative research program will enable Professor Raahemi to explore new and more efficient algorithms that will help us to better manage population health hazards, will make us more secure from online fraudsters, and will help organizations and business to take full potential of the Internet connectivity.