Hi, there! I'm Anmol.
I am a Master of Industrial Engineering graduate from The University of Toronto, Canada. Project Planning & Machine Learning Enthusiast with expertise in Data Science technologies. And I'm passionate about using data to make the world a better place! |
Stuff I'm Good At
Assorted Collection of the Finest Work I have done so far...
Neural Machine Translation from English to French
Neural Machine Translation exhibited using an Encoder-Decoder (LSTM) model integrating Bahdanau Attention Mechanism & Teacher Forcing through Tensorflow-Keras library. After training, presented inference through Greedy Search and Beam Search & model evaluation using BLEU metric
|
Comparison of CNN and RNN on sequence learning
Analyzed and Compared performance of 1-D CNN & RNN (Bi-LSTM) on NLP sequential learning using Tensorflow-Keras in a classification task to predict the publisher based on articles’ content. Showed that with careful architecture plan, CNN works in comparable terms with RNN (both achieving 87% accuracy)
|
Object Detection - YOLOv3
YOLOv3 State-of-the-art Object Detection algorithm is applied to localize prohibited items in baggage. Project aimed to automate this procedure alleviating human errors and associated costs.
|
Predicting Canadian Federal Elections 2019
Transfer learning from a generalized Logistic Regression model on twitter dataset to assess public sentiment towards all parties. Turns out Liberals would have won the elections... no surprise there...
|
Social Network Graph Analysis & Prediction
Building Social Network Graphs using Twitter Data and Predicting Links that could appear in the future. To compare, imagine forecasting who you're gonna tag in your next tweet...
|
Information Retrieval using Whoosh
Not a perfect SEO. But I will take an Rprec of 36.7% anytime of the day ;)
Given a corpus of documents, this will tell you how you can build an Information Retrieval model given a query to the model |
Some More for the explorers...
Recommendation Systems Algorithms Comparison
Basic concepts behind Recommendation Systems (cuz you wanna know how Netflix recommends you movies all the time)
|
Docker-Flask-MongoDB RESTFul APIs
Wanna make money out of your ML models? API is the way to go...
The three APIs deal with Image Classification, Text Similarity and a simple Bank Transactions app |
Miscellaneous Work
Association Rule Learning
So do you know why your grocery store puts bread next to milk? Click below to know why... |
Calgary Hotels Customers' Review Analysis from TripAdvisor
A deep analysis of hotel reviews in Calgary using Part-of-Speech and Regular Expressions combined. I bet you didn't know this before... |
|
|