Research

Database

• My undergraduate thesis is focused on evaluating performance of MOLAP (Multidimensional Online Analytical Processing) on EKA (Extendible Karnaugh Array) structure under supervision of Dr. K.M. Azharul Hasan.

• EKA is an efficient scheme which has better performance than other data structures that we have tested in our research.

• In our research work we used EKA as basic structure for implementing incremental aggregation of data and evaluate its performance over other approaches.

• We used Multidimensional Online Analytical Processing (MOLAP) which stores data in optimized multi-dimensional array storage, rather than in a relational database.

• We created MOLAP data cube using Traditional Multidimensional Array (TMA) and EKA scheme and compare incremental aggregation with Relational Online Analytical Processing (ROLAP).

• The effective outcome of EKA structure for incremental aggregation on MOLAP structure is shown by some experimental results.

• We published a conference paper on this research, Which can be seen here: Publications

• We are now thinking about extending our work to publish a Journal.

Data Mining

• While working with database and data wire-housing, I read the book ''Data Mining: Concepts and Techniques'' by Jiawei Han, Micheline Kamber and Jian Pei.

• Then I knew about data visualization, data storage, classification clustering etc.

• Interest in Data Mining has grown rapidly in my mind.

• Also studied and implemented recommendation system, clustering algorithms in python.

Machine Learning

• As a subdivision of Data Mining, Machine learning is a great topic of interest to me.

• I have completed a course in Coursera named ''Machine Learning" by Andrew Ng, that can be seen here: Online Courses

• We (me and my colleagues) have published two conference papers in this area using Machine Learning Techniques.

• One publication is about planning a itinerary plan for tourist using k-means clustering, TSP and 2-opt algorithms. We used python here.

• Another is about predicting crime trend of Bangladesh using linear regression. We used MATLAB for this.

• Both of the publications can be found here: Publications

• We have plan to visualize crime trends with other Machine Learning Techniques.

• Now I am learning Sparse Auto Encoder, Softmax Function and Convolutional Neural Network from ''Unsupervised Feature Learning. and Deep Learning'' by Stanford University. I will apply those techniques to recognize bangla letters and digits.

Data Science

• I have completed two data science courses Coursera which can be found here: Online Courses

• Hope to finish "Data Science Specialization" by John Hopkins University.

• Have a deep interest in this area.

Image Processing

• To apply Machine Learning algorithms on Image Dataset, Preprocessing and cleaning is necessary. For this purpose I am reading a book named "Digital Image Processing Using Matlab" by Rafael C. Gonzalez, Richard E. Woods and Steven L. Eddins.

Networking

• As a CCNA instructor of CISCO Local Academy, KUET, I am greatly introduced with networking world.

• To learn about new networking technologies and protocols and completed a masters level course named "Advanced Networking" Where I learn about Vehicular Ad-Hoc Network (VANET), Mobile Ad-Hoc Network (MANET) and various protocols used in these systems.