The following forefront areas have been identified to carry out the research studies.
- Advanced databases
- Artificial intelligence
- Parallel processing
- Software engineering
- Wi-Fi and blue tooth technology
Current Research topics
- Data Mining
- Soft Computing
- Computer Vision
- Wireless Sensor Network
Research Project funded by Savitribai Phule Pune University.
- Design of Classifier for Detection of Diabetes using Soft Computing
The usage of data mining techniques in clinical medicine datasets has opened new avenues in medical profession and soft computing. The motivation for this thesis arose as a consequence of studying new applications of predictive data mining methodologies to clinical data and the need of real users using these applications. This research work has been done to implement and evaluate a predictive model (classifier) for clinical data related to esp. diabetic patients.
Other Research Projects
- The development of Video Surveillance using IoT to secure the bank
Under this research project, the video surveillance application for banking security is being developed using IoT. In this project, the actual image of moving object is captured by CCT camera and, then the image is processed by using image processing techniques. If the moving object found live, then the system is capable to send video to concern authority to signal the misbehavior in bank or any financial organization.
- The development and implementation of fast algorithm for vision – based hand gesture recognition
The development of ubiquitous computing has made computer very important in our daily life. The development of user interfaces forms the change in the human computer interaction (HCI) devices. The aim of this research is to build up a real-time method which is able to accept instructions given by means of hand gestures to expand the ways that people can interact with their computers for playing video, enhancing the video quality and replacement of objects in a video frames by using hand gestures .The enhancement in fast algorithm is the major feature of the research.\
- Wireless Security & Privacy
The Security and Privacy of computer has high-impact, innovative, and exciting computer security research. Our current research interests include security and privacy for Wireless LAN 802.11a/b/g, mobile device security.
- Variance analysis in software fault prediction model using support vector machine
In this research machine learning method is used for predicting fault, i.e. support vector machine to predict the accuracy of the model predicted. The proposed models are validated using dataset collected from Open Source software. The results are analyzed using Area under the Curve (AUC) obtained from Receiver Operating Characteristics (ROC) analysis. The results show that the model predicted using the support vector machine outperformed all the other models. Hence, based on these results it is reasonable to claim that quality models have a significant relevance with object oriented metrics and that machine learning methods have a comparable performance with supervised learning methods.
- Categorization of the Documents using Machine Learning Techniques
Clustering is nothing but grouping similar items together. Clustering documents can be very useful for Information retrieval, News papers publishing house, Government agencies etc. K-means is simple to implement, widely used clustering algorithm used for clustering. But to process large documents using one machine is good solution. So Apache Hadoop which supports Mapreduce framework can be a good solution. Hadoop uses distributed computing and distributes task to multiple nodes (map stage) and then combine all outputs together (reduce stage) to generate final output. All these tasks executed in parallel. So performance can be increased. Checking cluster quality is also equally important. For that Davies Bouldin Index is used.
- DSP based vector controlled 3 phase Induction Motor
AC induction motors, which contain a cage, are very popular in variable-speed drives. They are simple, rugged, inexpensive and available at all power ratings. Progress in the field of power electronics and microelectronics enables the application of induction motors for high-performance drives, where traditionally only DC motors were applied. This drive application allows vector control of the AC induction motor using Digital Signal Processor (DSP) running in a closed-speed loop with the speed / position sensor coupled to the shaft.
- A Survey on Examination Scheduling Problem (ESP) and Hyper-Heuristic approaches for solving ESP.
Examination Scheduling is a divergent, combinatorial optimization problem which tends to be solved using stochastic search approaches such as evolutionary algorithms (EAs) and heuristic methods such as Hill-climbing, Simulated annealing, Tabu search, Genetic algorithms, Graph coloring etc. In this research on examination Scheduling Problem (ESP) and Hyper-Heuristics approaches used to solve Examination Scheduling.
- Reliability Enhancement in WSN Using Loss Recovery Model
Wireless sensor network (WSN) is a network of spatially distributed sensors communicating each other. They are transmitting their data to central node i.e. sink. For any network it should be reliable and energy efficient in nature. In multi-hop communication data flows through several hops but 100% data is not transmitted to sink. Sometime the data rate is also high. Traffic load in multi-hop wireless sensor network is not distributed uniformly over the nodes so there are more chances of data loss. To prevent this loss a necessary actions should be taken. In loss recovery algorithm caching method is used to recover from loss of packets and to improve reliability of network.