*CFU - Crimean Federal University.
1. apriori - the intherpritation of APRIORI method. At the input, transactions with elements are given, the output is the last gluing of transaction elements with support greater than 2.
2. bulding-recomendation - recommendation system based on Cosine Similarity Algorithm. The object with the names of users and their ratings is sent to the input, the recommendation for a specific user is output.
3. kohonen-map-clustering - the implementation of the Kohonen self-organizing map for the clustering problem. The training set, the pace of training and the number of clusters are served at the input. An array of weights is formed independently. It can be used for real numbers as well as for binary ones.
4. neuron - neuron learning for the task of classification by state. At the input comes only a training set. The pace of learning and an array of weights are formed independently.
5. neural-network - Neural network for two neurons. It has input, hidden and output parts.
6. hopfield-network - Hopfield neural network. Work includes unidirectional and bidirectional network. At the entrance comes a set of samples and a sample for identification. After the learning process, a sample is identified.
7. hierarchical-clustering - Hierarchical clustering. The algorithm involves the separation of objects into clusters by constructing a graph. At the entrance dots arrive. At the output distribution on the clusters.
8. page-rank - Page-Rank algorithm. The algorithm involves the calculation of the ranks of web resources by the number of incoming and outgoing links. The object with the number of incoming and outgoing links on a specific web resource is sent to the input. At the output ranks of each of the web resources. In general, it is solved by the Gauss method.
A. Petrushin. 2016