Alishan Anwar

aspiring Data Scientist

"A machine learning enthusiast constantly learning, and finding meaning in life through patterns, one dataset at a time."

Twitter Sentiment Analysis - Spark Streaming

Developed an NLP Project of taking a keyword and retrieving latest tweets containing the keyword through Spark Streaming and analysing sentiment of the tweets with visualisations using neural networks (RNN) and running on a Flask web application.

Live demo available at Twitter Sentiment Analysis

Image courtesy: https://bit.ly/3HY4PU4

Indoor Scene Recognition

Created a deep learning model that predicts 67 indoor scene labels with 70% accuracy. Original data set of 15,000 images could only achieve 15-20% accuracy and it was improved by using augmented data of over 81,000 images on TensorFlow.

Live Demo - Upload any image and check at Indoor Identifier

Image courtesy: https://bit.ly/3HUU8l0

Backorder Prediction

A product being backordered means placing an order for a later delivery date after the product has sold out. It is a frustration experience for everyone involved in the supply chain i.e. buyers, sellers, distributors. If only there was a way to predict which products are likely to get backordered. this ML project is made with the objective of solving this problem by making this prediction for you using the product's performance forecast and supply chain variables.

Live demo available at Backorder Predictor

Image courtesy: https://bit.ly/3qihpHC

Hate Speech Detection

In a world unbridled with hatred, only a machine learning model can be objective enough to show us a mirror to our words. With that foresight, I developed a Logistic Regression ML model that detects of racist, sexist and in general hateful content in tweets.

Github Repository Link

Image courtesy: https://shutr.bz/3I09zIv