Title: User Interaction Analysis to Recommend Suitable Jobs in Career Oriented Social Networking Sites
Description: Actually this is a challenge of RECSYS 2016. We propose an easy algorithm that recommends job for a user using past dat. We used both collaborative filtering & content based technique to do this. Deploy our algorithm in Java.
Title: Development of National Pathological Data Warehouse (NPDW) for Bangladesh
Description: I worked on the development of National Health Data Warehouse of Bangladesh. I synthesized the data, preprocess the data to remove noise, normalize the data, design data warehouse for faster query processing.
Title: Computing Double Cut and Join (DCJ) Distance for Genomes with Duplicate Genes
Description: Propose a system to Compute Double Cut and Join (DCJ) Distance for Genomes with Duplicate Genes using idea from A* search. We try to develop a heuristic that will facilitate a search similar to A* to get DCJ with duplicate gene.
Title: Analyse spatial data for improving street food quality
Description: I developed an Android application named FoodGraber to collect street food location information. I have designed and developed the android application.
Here I listed My undergraduate research works:
We worked on predicting the information data spread center over a network in which the spread of information follows the popular Susceptible-Infected (SI) model based on geographic location. We use clustering approach to train a machine learning system that we use to predict new information spread center. I got well acquainted with the concepts, possibilities, requirements, and limitations of various machine learning algorithms.
I helped my project supervisor to build a BNGLA CORPUS, which is the largest corpus on Bengali language in Bangladesh. I collected sentences from the Daily Prothom Alo - a daily newspaper of Bangladesh, then format and categorize the words using python language. I do the following task
Predicting the success of marketing campaign of Banks for direct marketing (i.e. subscribe a deposit) based on Naïve bias classification as a part of ML lab.