An artificial neural network (ANN) is an interconnected group of artificial neurons that uses a mathematical or computational model for information processing based on a connectionist approach to computation. In most cases an ANN is an adaptive system that changes its structure based on external or internal information that flows through the network. In more practical terms neural networks are non-linear statistical data modeling tools. They can be used to model complex relationships between inputs and outputs or to find patterns in data. They can also be used for real life applications like data processing, including filtering, clustering, blind source separation and compression.
Utilizing these properties it is possible to work towards the development of algorithms for audio and image separation. These algorithms are capable of separating the mixed signals without prior information of their statistics and characteristics. If audio and image mixtures are successfully separated then these may help in solving criminal issues.
Keeping in mind highly congested networks and long distance links we are also planning to develop a group of data compression techniques that can be applied to images, scanned documents and videos; and hence, creating a system that adapts itself to the quality of service (QoS) offered by the Internet (Network) connection instead of expecting a specific QoS of the network.
Modeling, classification and fault detection of sensors using intelligent methods is another major research area. Gyroscopes and Accelerometers are the typical inertial sensors used by ISRO for motion analysis of the moving body during flights. The sensors deployed for on-board operations are selected after a series of experimentations and careful fault analysis. However, these experiments invariably follow a deterministic model for characterizing the sensor behavior and then classifying them into good and bad ones. These models are often not accurate enough to capture the system response and fail to handle uncertainty in operating conditions.
Committed to stronger relations between industry and academics, IITK is providing services to industry and R&D Institutions including Indian Space Research Organization, Bhabha Atomic Research Centre, Engineers India Limited, Indian Oil Corporation Limited, Steel Authority of India Limited, Aptech, CMC, UPERC, CRISIL, Deloitte Touche Tohmatsu, Seedata Technologies and IDFC.
Knowledge Solutions provider, KARMAA (Knowledge Acquisition Retention Management Assimilation and Application) is a suite of algorithms that mines information from databases using high-end techniques. It is available for a number of applications. The suite includes proven algorithms from Computer Science, Artificial Intelligence, Machine Learning, Neural Networks, Fuzzy Systems and Electrical Engineering. Most of the present areas have a plenty of data and the important issues to use them so that the best and most useful information can be arrived on. Through KARMAA, the industries are provided with useful methods and a powerful modeling environment for catering to the needs without having to understand the back-end technology.
Professor P. K. Kalra
Department of Electrical Engineering
Indian Institute of Technology Kanpur