Słowa kluczowe: sensors, information quality, railway data, spatial confirmation, adversarial samples, convolutional neural networks, classification, natural language processing, natural language inference, representation learning, word embeddings, machine learning, deep learning., entopry, stiching, optimization, reinforcement learning, value predictors, exploration, temporal expressions, recognition, TimeML, CRF, LSTM, BiLSTM, KGR10, FastText, Generative model, AutoEncoder, Wasserstein distances, Goodness of fit test (for normality), optimal transport distance, Wasserstein distance, autoencoder-based generative model., gradient descent, optimization methods, adaptive step size, dynamic learning rate, neural networks, neural network, generalization, gradient regularization, spectral norm, Frobenius norm, traffic optimization, metamodels, activation functions, genetic algorithm, cutting & packing, object packing problem, phi-functions, deep learning