Bhargav Sagiraju


Harnessing data towards a smarter future.

About Me

pip install earth Successfully installed earth from earth import People bhargav = People(firstname="Bhargav", lastname="Sagiraju") bhargav.location "Singapore" bhargav.education "MSc. Business Analytics, National University of Singapore" bhargav.skills ["Python", "R", "Flask", "Keras", "Scikit-learn", "Caret", "Pandas", "dplyr", "Matplotlib", "ggplot", "git"] bhargav.interests ["business-analytics", "research", "data-science", "science-fiction", "guitar", "problem-solving"]

Projects

Magnum-Opus

A CNN based Image Classifier used to classify Aadhaar Card, PAN Card and any other document using VGG16.

Keras Image Classification CNN

Source

Heat-Seeker

A Named Entity Recognition algorithm with a companion REST API to detect entities of interest in a SMS.

Flask spaCy NLP Named Entity Recognition

Source

Emotion Perception using Reinforcement Learning

Interactive UI offers choice between facial emotion detection through Raspberry Pi Camera and text emotion detection through an external Kernel using Python. The project won the Best Engineering Project Award for 2019 from Visvesvaraya Technological University.

Keras LSTM NLP OpenCV Raspberry Pi

Source

SPIRE

Digitalization of customer risk identification process with an XGBoost model to pick lifestyle indicators and assess customer risk based on installed apps on a customer's smartphone. The model provided a whole new way of assessing app data (package info).

dplyr xgboost

Source

K-prototypes

An experiment with K-prototypes algorithm with Kmodes library which clusters over a dataset with mixed feature types. The cluster-based customer segmentation helped create new sales strategies.

Pandas NumPy KModes

Source

Web Scraper

BeautifulSoup based web & notification scrapers to scrape exam results from the University (VTU) results website. The scrapers saved a lot of faculty's time & money spent on extracting student results.

BeautifulSoup urllib

Source

Coursera Capstone

A repo for the IBM Data Science Professional Certificate course on Coursera. I created a dashboard using Dash to visualize city data from London and Paris. I created a KMeans model to cluster similar neighbourhoods.

Pandas NumPy Dash

Source

IBM Advanced Data Science Capstone

A repo for the IBM Advanced Data Science Course. I modeled customer subscription preference with Apache Spark. The exercise was based on the telemarketing dataset obtained from the UCI Machine Learning repository.

Apache Spark Pandas Numpy

Source

Contact

Email  |  LinkedIn