Big data and deep learning

Traditional data processing algorithms are usually not capable to process big data. As matter of fact, usually big data is being defined as such which cannot be processed with traditional techniques. At the same time a progress of technology makes that humans are now overwhelm by big data. One way of processing big data is to use deep neural networks, which are difficult to train so often a combination of several learning algorithms are used. This approach however requires a lot of human involvement in design of such systems for very specific cases. In this presentations and universal architectures and training methods are described which can handle very complex systems with minimal human interactions. Another important problem is a visualization of multidimensional data sets in 2 or 3 dimensions so humans can see them. A new very effective visualization algorithm is also presented. The third issue is the pattern clustering. Round clusters can be separated relatively easy. In this presentation new and very fast clustering of complex shape is also presented.