FindAPhD Weekly PhD Newsletter | JOIN NOW FindAPhD Weekly PhD Newsletter | JOIN NOW

Deep Learning for stock market predictions: Study of Fama's international efficiency


This project is no longer listed on and may not be available.

Click here to search for PhD studentship opportunities
  Dr V Schetinin  Applications accepted all year round  Self-Funded PhD Students Only

About the Project

A theory of Efficient Market Hypothesis (EMH) has been introduced by Fama to analyse financial markets. In particular the EMH theory has been proven in some real cases under different conditions including financial crises and frauds. The EMH is assumed to be tested by using models designed for predicting the financial time series. It is reasonable to assume that a prediction model built on retrospective data provides more reliable results of the EMH test. This assumption motivated practitioners to analyse Machine Learning (ML) methods suitable for building models with a high prediction performance. This project aims to find new insights in deep learning for building high-performance prediction models. A recent example of powerful DL strategy is the Long Short-Term Memory. Another DL strategy is based on the Group Method of Data Handling (GMDH) capable of building multilayer neural-network models of a near-optimal complexity on given data. In particular the developed GMDH-type neural network has outperformed the prediction models built by the conventional ML methods on the Warsaw Stock Exchange data. The complexity of the designed neural-network models is defined by the number of layers and connections between neurons. The models were compared in terms of prediction errors. Based on these experiments, the previous study has reported a significantly smaller prediction error of the proposed method than that of the conventional autoregressive and "shallow’’ neural-network models. This will allow decision-makers to obtain advantages from a designed Deep Learning method.

Research questions: (1) to extend the prior knowledge of neural-network model structure to build a near optimal model's structure on given data (2) to explore the prediction performance of conventional Machine Learning in comparison with the Deep Learning and GMDH-type neural networks.

The deadlines are as follows:

For March starters:

International applicants - 30th November 2021

UK nationals - 18th January 2022

For October starters:

International applicants - 30th June 2022

UK nationals - 5th August 2022


Publications: Sulaiman R.B., Schetinin V. (2021) Deep Neural-Network Prediction for Study of Informational Efficiency. In: Arai K. (eds) Intelligent Systems and Applications. IntelliSys 2021. Lecture Notes in Networks and Systems, vol 295. Springer, Cham.
Search Suggestions
Search suggestions

Based on your current searches we recommend the following search filters.

PhD saved successfully
View saved PhDs