Time series analysis via Residual Neural Networks

Machine Learning

Time series analysis using Residual neural networks. The main task is to explore how different ResNets can be combined to solve complex time series problems.

Master

Usually, Recurrent neural networks are used for time series analysis. Nevertheless, there exists also work for time series analysis that utilizes convolutional neural networks (CNN). ResNets being an advanced CNN is the natural next step to improve time series classification and prediction.

Goal

The goal is to research and develop new deep neural network architectures that utilize ResNets for the task. The main task is to explore and evaluate the capabilities of ResNet based time series analysis for classification or prediction. The architectures should be evaluated and verified using different real-world datasets (we have datasets from different domains from sport to medicine that can be used).

Learning outcome

Deep understanding of CNNs and ResNets
Working on a real-world application
Collaboration with researchers
Possibility to implement and research a novel approach

Qualifications

Python programming
Knowledge about deep learning is an advantage

Supervisors

Pål Halvorsen
Michael Riegler
Steven Hicks

References

He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. InProceedings of the IEEE conference on computer vision and pattern recognition 2016 (pp. 770-778).