Srabanti Guha

Boise, ID ***** 901-***-**** adyk26@r.postjobfree.com

https://www.linkedin.com/in/srabanti-guha-670836218/ EDUCATIONAL BACKGROUND

● Masters in Computer Science: Boise State University, Boise, ID, USA. CGPA: 3.79/4.00 Graduation Date: May 2023

Project Title: Explaining Misallocated Exposure across Multiple Rankings of Recommender System. Supervisor: Dr. Micheal Ekstrand

● Bachelor of Science in Computer Science and Engineering, Chittagong University of Engineering and Technology, Bangladesh

CGPA: 3.32/4.00 Graduation Date: December 2018

PERSONAL SKILLS

● Operating Systems: Windows, Linux

● Coding: Python, Java, C, PySpark, jQuery, HTML, CSS

● Technologies/Environment: MySQL

● Model Explanation Technique: LIME, SHAP, EXS, LIMERS

● Framework : Hadoop, Flask

KEY ACADEMIC PROJECTS

● Understanding global behavior of information retrieval model by model explanation technique Used information retrieval model (word2vec-based Vector Space Model) from genism python library to retrieve relevant documents in response to a set of queries, applied and analyzed model explanation technique “EXS” (Explainable search using local model agnostic interpretability) to access the global behavior of that information retrieval model.

● Develop a search engine that can retrieve relevant documents with snippets for a query and provides query suggestions

Created an index structure using Python for each word of the documents and retrieve top five relevant documents with their title and snippet (highly similar two sentences of the retrieved documents with the query).

● Modify the previous search engine as web application using flask, increase efficiency of retrieval and remove objectionable (unsuitable for kids) query suggestions Reorganized the indexing structure of words by dividing it into distinct blocks and incorporated alterations to the coding structure in order to increase the speed of retrieving. Removed bad query suggestions by using readily available bad word lists from Google and Facebook. Also developed user interface by HTML and CSS.

● Fake news detection by machine learning algorithms Extracted writing style and emotion-based features and calculate readability scores for fake and real news. Used those features with binary classification models such as Logistic Regression, SVC, Random Forest, Catboost, XGboost, and Extra Trees to detect fake news.

● Develop a database and supporting application for a Library logical data model Developed database in MYSQL and Library application in Python to insert, update and delete Books of different library branches for students. Also, included database trigger in the database to keep track of time, date of insertion, updating and deletion.

● Develop end-to-end machine learning project on (a) flight fare prediction (b) car price prediction with Python web framework flask.

Performed exploratory data analysis and feature selection. Also, used HTML and CSS for user interface design and flask for web app.

COURSES

Introduction to Data Science, Machine Learning, Databases, Large Scale Data Analysis, Information Retrieval, Advance Information Retrieval, Social Media Mining, Design and Analysis of Algorithm. WORK EXPERIENCE

Industrial trainee: Grameenphone (leading telecom operator in Bangladesh), April 2017 - May 2017. Practical experience in cloud computing, telecommunication, and networking using 3G and 4G technology. Teaching Assistant: Works as TA in “Programming Languages” in Boise State University. Responsible for grading, tutoring in a class size of 30 students.

Work Authorization

Lawful permanent resident

Contact this candidate