This project focus on predicting the hourly demand for demand rentals and returns at each station of the BIXI's bike sharing system. Our proposed model uses temporal and weather features to predict the main characteristics of the traffic demand (e.g. mean and variance of trips from/to each station). The model first extracts the main traffic behaviors from the bike stations. These simplified behaviors are then predicted and used to perform station-level predictions based on machine learning and statistical inference techniques. Finally, the model determines inventory intervals, which are often used by bike sharing companies for their online rebalancing operations.

The communication and computing infrastructure has evolved through the years, getting more efficient, sophisticated, integrated and networked. Newer mobile devices (including smart robots or autonomous cars) and servers often contain 8 or more cores in their central processing unit. These systems are based on heterogeneous processors, with efficient traditional central processing units, but also with co-processing units optimised for graphics (GPGPUs with thousands of cores), networking, signal processing or even for Machine Learning. These co-processing units are highly parallel and often contain over 8 billion logic elements (transistors) each. Adding to this complexity is the increasing reliance on virtualisation, which hides the specificities of the hardware, allowing an application to run on several different processor models, but makes the performance more difficult to analyse. As a result, even a simple operation such as initiating a phone call, making a Web search, routing a packet or displaying a video frame, can involve many parallel cores on more than one processing unit, possibly on several servers. Moreover, the same operation, a few seconds later, may be served in a different way by different cores and physical servers. Therefore, understanding the performance of these operations has become extremely difficult and the tools for that purpose are severely lacking. In this project, the tracing, monitoring, profiling and debugging tools for manycore systems will be extended to efficiently extract information from all units in all layers, from the hardware to the application, and cope with the large number (several thousands) of cores. Furthermore, new methods and algorithms will be developed to automate the analysis of the extracted monitoring data. As a result, the designers and operators of distributed applications on mobile devices, cloud servers and other heterogeneous computing systems, will have the tools in hand to quickly analyse their system performance, automatically or manually find problems, and optimise operations.

Our current speed of data generation combined with storage capacity increases have given rise to new paradigms in computing. For example, according to the IBM's website, there are approximately 695,000 status updates and 11 million instant messages sent every minute on Facebook. However, many organizations have faced the problem of having a lot of data, but poor knowledge about them. Clustering methods help to automatically identify unobserved groups for a set of data objects, and are currently being radically transformed by the size, the variety and the nature of the available data, i.e., by so-called Big Data. This research program focuses on the development of scalable algorithms for Big data clustering. This will be achieved both by: (i) redesigning well-known successful serial algorithms for scalability, leveraging their main theoretical ideas; and (ii) developing new algorithms and heuristics using new programming paradigms associated with Big Data. In summary, the objectives of this research program are : A. Produce an extensive survey of the existing Big Data clustering methods in order to provide a complete panorama about which ones can be adapted to approach Big data. B. Redesign successful serial clustering algorithms, leveraging their main theoretical ideas to work with the new programming paradigms and computational tools from Big Data. C. Develop algorithms for semi-supervised Big Data clustering, incorporating supplementary information provided by the user into the clustering decision process. D. Develop Big Data clustering algorithms for new emerging applications. E. Provide a repository of Big Data software and clustering algorithms with guaranteed effectiveness. The software and algorithms developed in this research program are expected to constitute new benchmarks to the field, allowing larger datasets to be tackled with efficiency and effectiveness, leading to new insights in commerce, industry and academia.Moreover, given the Big Data expert shortage in Canada and worldwide, this research program will help to form and train specialists who will be able to master the scientific and technological issues emerging from the Big Data explosion. In the long term, the findings of this research program will support data mining-driven decision making in the Internet of Things (IoT) era in which huge amounts of data are generated in real-time from the most varied sources and devices (e.g. vehicles, sensors, home appliances, etc.). By combining massive data processing, machine learning techniques and algorithms, IoT devices will be able to perform clustering as well as other classification tasks on a large scale.