Portfolio of my data science projects for academic & self-learning.

• Web-Crawling

  • Zomato Scrapper: Uses Scrappy Framework for collection of Menus, Reviews & Restaurant's information.

• Machine Learning

  • Sentiment Analysis of Restaurant Reviews: Built in Python. Cleaned and processed raw reviews for analysis of restaurant dishes using NLTK.

  • German Credit Risk To classifies people with Good or Bad Credit Risks. Built in R & Python. Implemented with Random Forest Algorithm.

  • Sonar Classification: Built in Python. To predict metal or rock objects from sonar return data. Implemeted and compared with SVM & Logistic Regression.

    • Dataset: https://archive.ics.uci.edu/ml/datasets/Connectionist+Bench+(Sonar,+Mines+vs.+Rocks)

• Data Mining

  • Data Mining Process (CRISP-DM) & Application: The PDF has 2 parts. Part A is a "Critique" on “The CRISP-DM Model: The New Blueprint for Data Mining” Volume 5 Number 4 Fall 2000 - Colin Shearer. Part B is a paper on "Smartphone Data Mining"

  • New York City Taxi Trip Duration: Prediction of New York City Taxi Trip Duration using CRISP-DM Methodology. Used Tableau for creating visualizations. Modelling is done using Rapid Miner's Auto Modelling feature. We find that gradient boositng algorithm works best and a GB ML Model is created. Model is then converted to PMML file for deployment.

    • Dataset: https://www.kaggle.com/c/nyc-taxi-trip-duration/data

• Data Visualization

  • Exploring Crime In Chicago (2001-2018)

  • Explanatory Analysis of Bank Churn