How to Predict Stock Prices with LSTM Model

Kannan (~kannan00)




Deep learning is best suited for sequential problems and financial time series being sequential in nature, deep learning shows promising results in financial time series prediction. This workshop will focus on the application of neural networks in finance and help you to get started on Deep Learning by applying the concepts to a real-world example.

The workshop will demystify the concepts by explaining the basic building blocks of neural networks and will introduce you to the popular Keras API framework. We will then see a real-world example of how this framework can be used in predicting stock prices using an LSTM model.

In the past few years, a lot of academic papers were published using neural networks to predict stock prices. Until recently, the ability to predict these models were restricted to academics. But, with libraries like Tensorflow and Keras, we can now build powerful predictive models.

The session will focus on the following agenda

  • Evaluation of Artificial Intelligence
  • Deep Neural Network framework
  • Building Blocks of Deep Learning
  • Overview of LSTM model
  • Introduction to TensorFlow, Keras
  • Data Retrieval and Preprocessing
  • Building an LSTM Model for stock price prediction

Setup for the workshop

  • Working Python environment (using anaconda distribution)
  • Installation of TensorFlow, Keras libraries

What will you gain from this workshop

  • Application of deep neural networks in finance
  • Hands-on experience in building LSTM model for price prediction


  • Python basics
  • Understanding of Machine / Deep Learning
  • Working Python environment (using anaconda distribution)
  • Installation of TensorFlow, Keras libraries

Video URL:

Content URLs:

Content Link

Speaker Info:

Kannan is a quant researcher and a data science enthusiast. He focuses on extracting information asymmetries and market inefficiencies through data analytics. During the past sixteen years, he headed a commodity research desk, oversaw a multi-asset structured products business and managed a multi-million dollar long-dated options book, among other roles.

He is a strong advocate of financial data science and was exposed to Bloomberg quant and data science platform - BQUANT - during his tenure at Bloomberg LP.

As a Pythonista, he has developed comprehensive ‘Python For Derivatives’ & ‘Python For Quant Finance’ modules, both of which are now part of the curriculum in reputed domestic and international institutions. He is a cohort of Paul Wilmott’s Certificate in Quantitative Finance and currently the President of CQF Mumbai Society.

His work and articles can be accessed through and

Speaker Links:

Section: Data Science, Machine Learning and AI
Type: Workshop
Target Audience: Intermediate
Last Updated: