Data Analytics
Predictive analytics, decision analytics, modeling, simulation, and optimization.
Data Analytics technologies aim to improve the performance of devices, systems, and business processes by means of collecting data, constructing predictive models from that data, and making improved decisions based on the constructed models. The Data Analytics group at MERL has been working on both predictive and decision analytics, as well as supporting fields such as signal processing, numerical methods, and information systems infrastructure.
The focus of the group is on innovative high-performance algorithms that can be applied to various product lines of Mitsubishi Electric, including electrical power systems, various transportation systems (trains, elevators, car navigation), heating, ventilation, and air conditioning (HVAC) systems and solutions, and factory automation. The application of these algorithms minimizes costs, maximizes profits, increases reliability, improves energy efficiency, and reduces environmental impact of products.
Quick Links
-
Research Areas
Predictive Modeling
Research on predictive analytics, supported by advances in the field of statistical machine learning, aims to create accurate data-driven models of electromechanical and thermodynamical systems, as well as models of complex natural and man-made phenomena, such as road traffic and demand for electrical power.
Decision Optimization
Numerical methods for fast solution of network problems find application in the analysis of electrical power systems and Smart Grids that include renewable power sources with intermittent output, as well as highly variable loads such as electrical vehicles. Many problems in transportation systems, such as train operation optimization, group elevator scheduling, and route guidance for car navigation, can be reduced to planning and optimization problems. Energy consumption in buildings can be minimized by solving sequential decision and optimization problems, both at the level of individual air conditioning devices, as well as at the level of entire buildings.
-
Researchers
-
Awards
-
AWARD DEIM 2010 Best Paper Award Date: February 1, 2010
Awarded to:
MERL Contact: William S. Yerazunis
Research Area: Data Analytics -
AWARD SICE 2008 Best Paper Award Date: July 1, 2008
Awarded to:
MERL Contact: Daniel N. Nikovski
Research Area: Data Analytics
See All Awards for MERL -
-
News & Events
-
NEWS MERL researchers presented 11 papers at ACC 2017 (American Controls Conference) Date: May 24, 2017 - May 26, 2017
MERL Contacts: Mouhacine Benosman; Daniel J. Burns; Claus Danielson; Stefano Di Cairano; Amir-massoud Farahmand; Abraham M. Goldsmith; Uros V. Kalabic; Andrew Knyazev; Saleh Nabi; Daniel N. Nikovski; Arvind U. Raghunathan; Yebin Wang
Research Areas: Data Analytics, Mechatronics, Algorithms, Advanced Control Systems, Dynamical Systems, Machine LearningBrief- Talks were presented by members of several groups at MERL and covered a wide range of topics:
- Similarity-Based Vehicle-Motion Prediction
- Transfer Operator Based Approach for Optimal Stabilization of Stochastic Systems
- Extended command governors for constraint enforcement in dual stage processing machines
- Cooperative Optimal Output Regulation of Multi-Agent Systems Using Adaptive Dynamic Programming
- Deep Reinforcement Learning for Partial Differential Equation Control
- Indirect Adaptive MPC for Output Tracking of Uncertain Linear Polytopic Systems
- Constraint Satisfaction for Switched Linear Systems with Restricted Dwell-Time
- Path Planning and Integrated Collision Avoidance for Autonomous Vehicles
- Least Squares Dynamics in Newton-Krylov Model Predictive Control
- A Neuro-Adaptive Architecture for Extremum Seeking Control Using Hybrid Learning Dynamics
- Robust POD Model Stabilization for the 3D Boussinesq Equations Based on Lyapunov Theory and Extremum Seeking
- Talks were presented by members of several groups at MERL and covered a wide range of topics:
-
NEWS MERL's breakthrough speech separation technology featured in Mitsubishi Electric Corporation's Annual R&D Open House Date: May 24, 2017
Where: Tokyo, Japan
MERL Contacts: Bret A. Harsham; John R. Hershey; Jonathan Le Roux
Research Areas: Multimedia, Speech & AudioBrief- Mitsubishi Electric Corporation announced that it has created the world's first technology that separates in real time the simultaneous speech of multiple unknown speakers recorded with a single microphone. It's a key step towards building machines that can interact in noisy environments, in the same way that humans can have meaningful conversations in the presence of many other conversations. In tests, the simultaneous speeches of two and three people were separated with up to 90 and 80 percent accuracy, respectively. The novel technology, which was realized with Mitsubishi Electric's proprietary "Deep Clustering" method based on artificial intelligence (AI), is expected to contribute to more intelligible voice communications and more accurate automatic speech recognition. A characteristic feature of this approach is its versatility, in the sense that voices can be separated regardless of their language or the gender of the speakers. A live speech separation demonstration that took place on May 24 in Tokyo, Japan, was widely covered by the Japanese media, with reports by three of the main Japanese TV stations and multiple articles in print and online newspapers. The technology is based on recent research by MERL's Speech and Audio team.
Links:
Mitsubishi Electric Corporation Press Release
MERL Deep Clustering Demo
Media Coverage:
Fuji TV, News, "Minna no Mirai" (Japanese)
The Nikkei (Japanese)
Nikkei Technology Online (Japanese)
Sankei Biz (Japanese)
EE Times Japan (Japanese)
ITpro (Japanese)
Nikkan Sports (Japanese)
Nikkan Kogyo Shimbun (Japanese)
Dempa Shimbun (Japanese)
Il Sole 24 Ore (Italian)
IEEE Spectrum (English)
- Mitsubishi Electric Corporation announced that it has created the world's first technology that separates in real time the simultaneous speech of multiple unknown speakers recorded with a single microphone. It's a key step towards building machines that can interact in noisy environments, in the same way that humans can have meaningful conversations in the presence of many other conversations. In tests, the simultaneous speeches of two and three people were separated with up to 90 and 80 percent accuracy, respectively. The novel technology, which was realized with Mitsubishi Electric's proprietary "Deep Clustering" method based on artificial intelligence (AI), is expected to contribute to more intelligible voice communications and more accurate automatic speech recognition. A characteristic feature of this approach is its versatility, in the sense that voices can be separated regardless of their language or the gender of the speakers. A live speech separation demonstration that took place on May 24 in Tokyo, Japan, was widely covered by the Japanese media, with reports by three of the main Japanese TV stations and multiple articles in print and online newspapers. The technology is based on recent research by MERL's Speech and Audio team.
See All News & Events for Data Analytics -
-
Recent Publications
- , "Solving Nearly-Separable Quadratic Optimization Problems as Nonsmooth Equations", Computational Optimization and Applications, DOI: 10.1007/s10589-017-9895-8, Vol. 67, No. 2, pp. 317-360, June 2017.
- , "Submodular Function Maximization for Group Elevator Scheduling", International Conference on Automated Planning and Scheduling (ICAPS), June 2017.
- , "A Novel BESS-based Fast Synchronization Method for Power Grids", IEEE PES PowerTech, June 2017.
- , "Towards Stability in Learning-based Control: A Bayesian Optimization-based Adaptive Controller", The Multi-disciplinary Conference on Reinforcement Learning and Decision Making, June 2017.
- , "Deep Reinforcement Learning for Partial Differential Equation Control", American Control Conference (ACC), May 2017.
- , "Transfer Operator-Based Approach for Optimal Stabilization of Stochastic Systems", American Control Conference (ACC), May 2017.
- , "Value-Aware Loss Function for Model Learning in Reinforcement Learning", Artificial Intelligence and Statistics (AISTATS), April 2017.
- , "Attentional Network for Visual Object Detection", arXiv, January 2017.BibTeX Download PDFRead TR2017-028
- @techreport{MERL_TR2017-028,
- author = {Hara, K. and Liu, M.-Y. and Tuzel, C.O. and Farahmand, A.-M.},
- title = {Attentional Network for Visual Object Detection},
- institution = {MERL - Mitsubishi Electric Research Laboratories},
- address = {Cambridge, MA 02139},
- number = {TR2017-028},
- month = jan,
- year = 2017,
- url = {http://www.merl.com/publications/TR2017-028/}
- }
-
Videos
-
Free Downloads
