This folder contains some of my Machine Learning and Deep Learning projects.
The projects are as follows:
-
FB-Chatbot : I tried creating my personal FB Chatbot using my Facebook data. The data was downloaded in JSON format and then preprocessed in order to get all the messages and responses in one single file. Then a 3 layer LSTM model was created to train on this one huge message-response file. After training, created a web app using REST APIs and deployed it on Heroku (so that you can chat with me ARTIFICIALLY!!! :D )
-
Predicting energy consumption on RECS dataset : Here my main goal was to use three popular Regression algorithms and see which one works better for the given dataset. I ended up using Linear Regressor, Elastic Net Regressor and LightGBM Regressor. The dataset had to be preprocessed, some of the preprocessing steps taken were, imputing missing values and robust scaling.
-
Stock price analysis : The objective here is to see the correlation between Twitter activities of a company and their stock price.From Yahoo Finance Microsoft's stock price dataset was used to find the correlation. This was a course project and was done in a group :)