What we offer
In many disciplines a key element of scientific research is the successful collection and analysis of data. Although conventional statistical methods give a good insight into the data, complex patterns, for example, cannot be identified with these techniques.
The application of data science, in turn, offers this detailed insight into the data. Data science is a multidisciplinary field combining expertise in statistics, data analysis, machine learning, and computer science that allows critical information to be extracted from the data. Due to these powerful methods, many research areas already apply data science techniques for in-depth analysis of datasets, and for making predictions from new observations.
How can SIS help you with the application of data science for your research?
The SIS Data Science Support is aimed at all scientists from ETH Zürich who want to use data science for their research, but do not have the capacity or the required expertise to do so themselves. Our Data Science Support ranges from consulting regarding suitable approaches and methods (in cooperation with the Machine Learning Consulting Service ) to the implementation and provision of entire solutions. In addition, we also collaborate closely with the Swiss Data Science Center .
Possible fields of application include (but are not limited to):
- Numerical/categorical tabular data → machine learning methods for classic regression classification tasks;
- Image data → object detection, image and pattern recognition, 3D reconstruction;
- Time series → identification of complex patterns in time series and prediction of time series;
- Text data → text classification.
Selected projects
- Segmentation and analysis of microscopic images
- OCR (optical character recognition) and topic modelling on large text corpus
- Multi-omics analysis using regression and classification
- Gene sequence analysis
- Data collection & preparation
Courses
We also offer a two day workshop Introduction to Machine Learning using Python targeting scientists without previous machine learning experience.