Beyond the media buzz, big data is a major strategic topic at the heart of great economic and societal challenges. Its impact is now perceived in almost every field of human activity: scientific research, health, finance, construction, e-commerce, security, transportation, etc.
Technological advances, the ubiquity of sensors (embedded systems, the internet of things, the Web, etc.) and the boom of social networks come with a veritable flood of data, putting the information sciences at the center of the big data valorisation process. Beyond collection and storage, the challenge is to analyse the data in order to optimize decision making and build even more efficient new applications, such as predictive maintenance in networks and transportation, commercial targeting, biometry, etc.
The statistical processing of massive data brings together applied maths and computer science through a quickly expanding discipline: machine learning. The variety of available data (numbers, images, text, signals), their high dimensionality and volume often make traditional statistical methods, which rely on human pre-processing and long modelling work, totally ineffective. Machine learning aims to elaborate and study algorithms, usually with predictive goals, that enable machines to learn automatically from data and perform tasks in an efficient way (such as recommendation engines and anomaly detection). The Chair of “Machine Learning for Big Data” aims to produce methodological research that would respond to the challenge of the statistical analysis of big data and to drive the education in that field at Telecom ParisTech.
A skill in the heart of Télécom ParisTech
With high-level research and teaching, covering all the information technologies and their uses, Telecom ParisTech has established a unique innovation ecosystem based on a strong interaction between its education, its research center and the two start-up incubators. Its tight bonds with the industry make this school a privileged eyewitness to the rise of the big data phenomenon and its technological impact on the education and research fields. Learn more about this at www.telecom-paristech.fr/bigdata (in French).
One of the strengths of Telecom ParisTech has always been its skills and expertise regarding the processing of structured information (signal, networks, images, videos, textual data). The omnipresence of machine learning in the school’s research and development activities, combined with a capacity to adapt the analysis methods to the nature of data and to the information systems, enables Télécom ParisTech to significantly contribute to the state-of-the-art.
An industrial Chair of research and teaching
Created in September 2013 with the support of the Telecom Foundation, and funded by four companies: Safran, PSA Group, Criteo and BNP Paribas, the Chair of “Machine Learning for Big Data” is directed by the mathematician Stéphan Clémençon, Professor in the department of Signal and Image Processing at Telecom ParisTech.
The Chair offers five methodological research axes in the field of machine learning, exemplified in concrete industrial applications like commercial targeting, recommendation, risk management, complex systems monitoring:
- Axis 1 - Reinforcement learning and stochastic optimization/simulation
- Axis 2 - Graph-mining and social network analysis
- Axis 3 - Ranking and Anomaly Detection
- Axis 4 - Cloud Learning and Distributed Learning Algorithms
- Axis 5 - Large Dimension Learning and Time Series/Data Streams
The goal is to collaboratively carry out, in interaction with its partners, cutting edge research activities in machine learning and education programs in the field of data science:
- “Big Data: management and analysis” Post-Master’s Degree
- “Data Sciences” Research Master, with the Ecole Polytechnique
- “Data Scientist” Specialized Studies Certificate (CES)
- “The basics for Big Data” MOOC