Data Science Enthusiast
Data Science Enthusiast having passion for analyzing, dissecting, manipulating data and building statistical models. Looking for a good opportunity to work on challenging projects where I can use my knowledge as well as skills which would enable me as a professional to grow while fulfilling organizational goals.
Tata Consultancy Services,Mumbai
Experience in setting up BWise reporting environment, Data warehouse loads.
Providing technical support for application users and optimizing performance of environment.
• Communication with clients, business users and other teams.
• Fixing data flow and quality issues.
• Use of BIS tool, JMeter and APIs for data migration. Real Time Gun Detection Using Surveillance Camera
Problem Statement: Propose a method to detect gun and mask carried by a person in a low-resolution image such as CCTV image. Approach: Designed a multi class classifier to detect guns and person wearing mask using faster RCNN-Inception-V2-COCO model and send alert messages with GPS details to nearby police station and stores. Tools Used: Python(Tensorflow,OpenCv)
Data Identity Hackathon
Problem Statement: To predict the performance of trainees (Pass/Fail) of a bank, given the demographic information and training program/test details. Approach: Conducted EDA with univariate, bivariate and multivariate analysis to find important attributes followed by missing value imputation. Built various classifier models such as Decision Tree, Logistic Regression and Random Forest. Selected the optimum one based on AUCROC score. Tools Used: Python
LeaderBoard: Top 10 Percentile.
Sentiment Analysis of IMDB Movie Reviews
Problem Statement: Predict the Sentiment of reviews which is binary. Approach: Performed basic pre-processing steps like transforming reviews to lowercase, removing markup, punctuations, numbers, extra spaces, stop words and Lemmatization of words. Created TF-IDF vectorizer and built various predictive models like Logistic Regression, Random Forest and Naïve Bayes.
Tools Used: Python
2018-01 - 2018-10
Praxis Business School
Post Graduate in Business Analytics
2011-06 - 2015-06
Amrita School of
B.Tech in Electrical and Electronics
Big Mart Sales Prediction
Problem Statement: To predict the sales for 1559 products across 10 stores in different cities using Predictive Modelling. Approach: Performed EDA, dealt with incorrect entries and imputed missing values in the data. Modified few categories and added new features so that much meaningful patterns can be learnt from data. Built different regression models using tuned parameters and made predictions. Tools Used: Python
Reducing the size of the image files using PCA
Problem Statement: Image files occupy a lot of space; our aim is to reduce the image size without reducing the clarity of the image. Approach: Created function where the user would input the image clarity and the file location. The function will reduce the file size by a PCA dimensionality reduction where the number of components will be based on the user input clarity. Additional arguments to replace the old images is also provided in the function.
Tools Used: R