Google Stock Price Time Series Prediction with RNN(LSTM) using pytorch from Scratch
Kaustabh Ganguly (~KaustabhGanguly) |
Description:
I will show you how to predict google stock price with the help of Deep Learning and Data Science . The predictions are not realistic as stock prices are very stochastic in nature and it's not possible till now to accurately predict it . As I'll only have 30 mins to talk , I can't train the data and show you as it'll take several hours for the model to train on google collab .
So , I will show you :
- Basics of Recurrent Neural Networks and LSTM
- Basics of pytorch
- Coding line by line with describing every words
- Then starting to train the model and prematurely closing it and move forward to show you the results that I'll bring with me after training .
Prerequisites:
You should have basic pyTorch understanding but I'll guide you anyways through the basics . Basic understanding of LSTM or RNN is preferred but not required .
Content URLs:
A sample code can be found here : https://github.com/KaustabhGanguly/Recurrent-Neural-Networks-to-predict-Google-Stock-Price
Speaker Info:
I'm a 3rd year Engineering student at Kalyani Government Engineering College , West Bengal , India . I'm a data science enthusiast and I interned at a machine learning startup called param.ai from June till August , 2018 . I'm obsessed with AI and my hobby is to study new cutting edge deep learning algorithms and research papers .
Speaker Links:
Follow me on github : github.com/kaustabhganguly Connect with me on linkedin : linkedin.com/in/kaustabh