Prof. Sowmya Kamath , Manifold Ranking (MR) is used to discover intrinsic geometrical structure of the IRMA medical data-set. As MR is computationally expensive, EMR is developed by building an anchor graph on the database instead of a traditional k-nearest neighbor graph and relevance feedback is maintained to improve ranking results. --Python & Matlab

Prof. Sowmya Kamath , Based on a cube model with SD-HCF for WS recommendation to consumers and ICF-UCF for potential consumer suggestion to providers. Involves significant concepts of NLP in creation of WS reviews data and applying sentiment analysis along with deep learning to assign rating which is integrated with predicted QoS value to obtain QoE. --Python-NLTK

Prof. Jaidhar C.D. , Extraction of various features from the metadata, API calls and context flow graphs of each application and training a SVM in an offline manner in order to leverage the higher computing power of a server. --Python-Androguard

Prof. Nagamma Patil , Robust TWSVM built to handle dataset with noise using second order cone programming formulations for classification. Improvement in computation time and classification accuracy over traditional SVM. Degenerating Robust TWSVM to overcome shortcomings of TWSVM. --Python & R

Prof. G.M. Reddy and Prof. Dinesh Naik , A toolbox that allows users to type through musical notes based buttons which represent the Braille language keys. It converts Braille images to the corresponding English text and vice versa through image processing, segmentation process and many other ML techniques like regression algorithms. --Python-OpenCV

Prof. Dinesh Naik , Web scraped data from various e-commerce sites and built an Item Based Collaborative Filter (IBCF). Implemented several similarity algorithms like correlation with weighted prediction technique to generate recommendations. Developed RDF Schema to obtain semantic based search with TF-IDF relevance and binary index tree storage with greedy DFS retrieval to reduce computation cost. --HTML, CSS3, PHP, RDF, Protege, OWL, MySQL, Visual Studio, Python & Selenium

Prof. Geetha V. , Implemented the Suzuki and Two Pass algorithms for CCL to identify objects and background in an image. Achieved thread level parallelism and run distributively using OpenMP and Open MPI directives. --C/C++, OpenMP & Open MPI

Mentored juniors and worked with them on applying data analysis and ML clustering algorithms on accident prone areas data of a region. The application captures location of user (if location belongs to danger or safe zone cluster) and alerts the driver of type of risks based on age, expertise and gender details. --Python-Django

Prof. Dinesh Naik , Web scraped data from various e-commerce sites and built an Item Based Collaborative Filter (IBCF). Implemented several similarity algorithms like correlation with weighted prediction technique to generate recommendations. Developed RDF Schema to obtain semantic based search with TF-IDF relevance and binary index tree storage with greedy DFS retrieval to reduce computation cost. --HTML, CSS3, PHP, RDF, Protege, OWL, MySQL, Visual Studio, Python & Selenium

Mr. Prakash P., Trained data containing multiple emotions and detected emotions like sadness and happiness by implementing a regression algorithm. Optimized code for detecting multiple faces at angles and incomplete faces. --Python and OpenCV

Miss Kajal Bangera, The GUI is implemented using HTML and CSS. The game’s algorithm and computations are performed using JavaScript. Developed new Finite State Automata for the heuristic of the game and compared it with the existing backtracking tree algorithm approach.

Implemented a user interface for the scrabble board, validated the word formed by the user, calculated score based on the word (letter and grid weightage) and realized the end of the game. Implemented an algorithm for AI. --Python

Used Java GUI framework. Worked with serialization and swings to store user data, check in and check out dates using a calendar and accordingly book a room based on it’s availability. --Java