Md. Yasin Kabir
Ph.D. Student, Computer Science.
Programming Language
Python (Most comfortable), C++, Java.
Machine Learning Library
PyTorch, Keras, TensorFlow.
Website Development
WordPress, PHP, HTML5, CSS3.
Data Visualization
D3.JS, Plotly.
Kaggle Competitions

I am Md. Yasin Kabira computer professional from Bangladesh. Currently, I am pursuing my Ph.D. in computer science at Missouri University of Science & Technology, Rolla, Missouri, USA. I am a Machine Learning(ML) enthusiast. Hence, my research and hobby both are ML focused. 

Computer Science, Missouri University of Science and Technology, USA.
Ph.D. Student2017 - Present
Computer Science and Telecommunication Engineering, Noakhali Science and Technology University, Bangladesh.
B.Sc. Engineering2009-2014
Department of Computer Science, Missouri University of Science and Technology, Rolla, Missouri, USA.
Graduate Research AssistantAug 2017 - Present
Verisign Labs, Verisign, Inc. Reston, Virginia, USA.
Graduate Research InternJun 2018 – Aug 2018
Computer Science and Telecommunication Engineering Noakhali Science and Technology University, Bangladesh.
Lecturer (On Leave)Nov 2015 – Jan 2017
Infosys Ltd., Mysore, India.
Software Development TraineeJan 2015 – Mar 2015
Current Research Projects
EMOCOV: Machine Learning for Emotion Detection

The primary purpose of this research work is to develop machine learning models to identify the emotions in the tweets and extract phrases from tweets for the specified emotions. I have started to Track, process, and analyze the tweets on COVID-19 since March 2020. With the help of some undergraduate students, we were able to manually annotate 10K tweets emotion. Further machine learning models were developed to achieve the primary goals. A basic real-time COVID-19 tweets tracking and data visualization are available at CoronaVis (mykabir.github.io). Currently, I am conducting an experimental analysis on the historical COVID-19 tweets that originated from the USA.  

Social media analysis for disaster management Using deep learning

The purpose of this project is to extract the information from social media and develop a system model that can help in the rescue, resource management, and recovery. Develop a deep learning model to classify the tweets and detect the tweets which seek helps during hurricane Harvey and Irma. Further, a rescue schedule algorithm is also developed to schedule the rescue operations of the identified people from the tweets. We have developed an application based on the research for real-world implementation, testing, and validation.

Key Kaggle Competition Projects
Tweet Sentiment ExtractionPosition: 13

This is an NLP problem. The task is to extract the support phrases for a specific sentiment label. Here labels can be neutral, positive, and negative. This is similar to question answering problems.

Bengali.AI Handwritten Grapheme ClassificationPosition: 14

Bengali is a complex language with enormous number of characters (grapheme roots, vowel and consonant diacritics, and graphemes). The aim of this competition is to classify a given image into respective characters. An image is a composite of one or multiple grapheme roots and diacritics. 

Understanding Clouds from Satellite ImagesPosition: 27

This is a competition for classifying four different clouds formation from satellite images. The goal is to use the cloud formation to determine the Earth’s climate.

Recent Publications