Subhankar Karmakar Portfolio

Data Analyst skilled in SQL, Tableau, Python, and MS Excel with research experience in the field of Biomedical Sciences and Engineering @Subhankar_linkedin

Loan Approval Prediction (In Python)

In this Machine Learning Project, I developed a model to predict loan approval based on gender, marital status etc. I performed EDA and pre-processing using Standard Scaler and then trained the model using Random Forest, Gaussian NB, Decision Tree, and Kneingbors. The highest accuracy achieved was 82.9 %. .

Credit Card Fraud Detection (Python)

In this Machine Learning Project, I analyzed the imbalanced credit card data from Kaggle and performed undersampling to evenly distribute the normal and fraud transactions. I split and trained the data in a Logistic Regression Model and then evaluated the model based on accuracy scores. .

Hotel Bookings Data Analysis Project (Python)

In this end-to-end Python project, I analyzed and visualised the Hotel Bookings dataset from Kaggle to investigate the reasons behind the high cancellation rates in City and Resort Hotels. I then prepared a report (pdf file) in which I discussed the findings and suggested solutions. I implemented libraries such as pandas, matplotlib, and seaborn for this project.

Currency Converter Chatbot (Python)

In this NLP (Natural Language Processing) project, I built a Dialogflow chatbot (named Converter Kaka) that can convert and tell the value of one currency to other and can also carry a small talk with the user. The chatbot was migrated to Telegram Messenger. Pycharm was used for the backend and libraries such as flask and requests were implemented.

Web Scraping Wikipedia Page (Python)

In this Web Scraping project, I scraped data of the largest Indian companies from Wikipedia, put it in a pandas dataframe and exported in a csv file. I used BeautifulSoup and pandas for this project.

Data Cleaning in SQL

In this project, I downloaded the dataset of Housing in Nashville, TN from github and simplified it by removing duplicates and unused columns, standardizing the data, and splitting columns into multiple individual columns that are more usable. I used statements such as CONVERT, ISNULL, PARSENAME, CASE, ROW_NUMBER, and PARTITION BY, and also created CTE.

Covid 19 Data Exploration
in SQL

In this project, I explored the dataset of COVID 19 and Covid Vaccinations by the World Bank. I used MS SQL Server and implemented JOINS, TEMP TABLES, and AGGREGATE Functions. I converted Data Types (using CONVERT and CAST statements) and created Views.