Excessive information noise and large multidimensionality of available data can impede comprehensive assessment of results, even by experts. Therefore, one of the basic tasks of modern data analysis is to extract the most important information contained in the available data set. By using advanced machine learning methods and Artificial Intelligence (AI), we can identify existing relationships and extract the most important information in a large data set within the context of the proposed research problem.
At QSAR Lab, we develop and use the latest technologies forming an essential part of an R&D project and can facilitate both scientific and business decision-making processes.
We offer use of advanced machine learning/artificial intelligence methods to retrieve relevant information contained in large data sets (e.g. chemical, medical).
Depending on the type of data and information that the client wants to obtain, a team of experts will create a flexible case study, covering all aspects of the study, such as: description of the research problem and hypotheses, description of the most appropriate and selected methods of machine learning for the proposed research problem, documentation containing reports with research results and their detailed description.
QSAR Lab experts conduct analysis using various techniques, including:
- supervised learning – methods based on input information and the end result and the search for relationships between them (regression and classification methods, such as linear and logistic regression, support vector machines, decision trees, Bayes classifiers, multilayer neural networks and image analysis)
- unsupervised learning – methods using only input information (without the end result), looking for dependencies, similarities, differences between the analyzed objects (e.g. cluster analysis, neural networks and SOM, analysis of main components, expectation-maximization algorithms).