Predictive analytics with data streams

Agenda:

6 pm: A talk by Dr Grazziela Figueredo

6:30 pm: Q&A, discussions and chat

Abstract:
Pre-processing, predictive and prescriptive analytics of data streams are areas of increasing research and commercial interest.

With the advent of sensor technology, high throughput data devices, web click streams, cloud computing and other technologies for constant big data gathering, there are several open research and business challenges to be addressed.

For instance, the dynamic nature of the data streams require algorithms capable of adapting over time to handle data drifts. Furthermore, the methodology employed to address the problem needs to be robust enough to effectively execute the processing and learn the high-speed arrival of data.

In this talk we discuss various aspects of data stream mining, focusing on the existing types of data streams, algorithms used based on traditional machine learning methods and their adaptation to concept drifts. We also discuss several opportunities and areas of development in streaming algorithms, which we are working at the moment.

We conclude our talk by presenting a practical application of a streaming algorithm developed to identify HGV incident hot spots, implemented in SCALA using Apache Spark.

Bio

GF is the Research and Innovation Leader and Senior Data Scientist at the Advanced Data Analysis Centre (ADAC) within the Digital Research Service at The University of Nottingham. The focus of her research is the development and application of techniques for systems simulation and intelligent data analysis. Her first PhD focused on developing an immune-inspired algorithm for instance selection in large data sets, at the Federal University of Rio de Janeiro. Her second PhD, completed at the University of Nottingham, focused on the translation of simulation approaches for immunology problems. She has been working with data analysis for a wide range of areas, including academic, medical and industrial clients. Within ADAC, her duties are to conduct independent research, to participate in collaborative research projects, to engage in business interactions, to secure internal and external funding, to supervise undergraduate and post graduate students, to participate in outreach activities and to disseminate scientific outcomes. As part of a more general ADAC remit, her job is to manage research projects and consultancy within the University of Nottingham and externally, with the mission to enhance current research and business by providing state-of-the-art tools and expertise in Data Science. In addition, she worked closely with several major companies, such as Ford, Microlise, Games Workshop and Unilever supporting data analytics teams, providing consultancy, promoting research knowledge transfer activities, as well as facilitating knowledge transfer partnerships (KTPs) and Innovate UK initiatives. Currently she is one of the academic advisors for a KTP with a company named PxTech in Derby. She was the academic researcher in the VEDAT Innovate UK project with Microlise (ref 101938).

We will be providing food (Pizza) and drinks which are kindly provided by our sponsor (RP Analytics).

Please don’t forget to RSVP so we have an idea about the numbers attending

See you all at University of Nottingham,

Data Science Nottingham team