An Ensemble Approach in Predictive Modeling of Big Data

Nowadays, the volume of information in all spheres is constantly increasing. After the scientists from Human Genome Project the first encoding of the human genome almost twenty years ago the number of studies on the sequences of the human genome is growing. However, despite this, there are many parts of the genome, which still require in-depth analysis. One of the directions in deciphering the human genome is DNA punctuation. DNA punctuation is the study of DNA secondary structures. Whole-genome experimentally confirmed annotation with DNA secondary structures still does not exist but they play an important role in the functioning of the human body. The present master thesis reveals application of ensemble methods to unpredictable patterns in the human genome and more specifically apply ensemble approach for classification genome sequences that form DNA secondary structures by belonging to one or the other genetic functional class.