Qualicision in Connection with Further Methods of Artificial Intelligence
Gradient Boosting – Advanced Form of Decision Trees...
An ensemble of trees is iteratively trained such that each successive tree tries to minimize the weaknesses of the previous tree, thus improving the overall result.
Convolutional neural networks – a characteristic of deep neural networks...
… which is particularly suitable for processing and classifying 2D and 3D input data, e.g. images and 3D models.
Natural Language Processing (NLP) – processing of natural language......
… with AI methods, e.g. for the detection of duplicates or the intension of texts.
Encoding and Decoding with Autoencoders...
An AI method for dimension reduction, anomaly detection or data synthesis.
Deep neural networks – an extension of neural networks...
… to extract prior data characteristics by interlacing of neural layers.
Node embedding to analyze delays in traffic networks
Neural networks – a method of machine learning...
… to process complex structural data as well as to realize classification, description and recognition.
Recurrent neural network – a characteristic of deep neural networks...
… which is particularly suitable for processing and classifying sequential input data, e.g. language and time series.
Machine learning – a method of information technology...
… which is often used as a synonym to the expression artifical intelligence to describe the prozess of automated data processing from the fields of big data, planning and intelligent assistence systems.
Supervised vs. unsupervised learning – are two methods of machine learning...
… that on the one hand training data with fixed results is available to process unknown samples (supervised). On the other hand the determination of unknown structures in data without availability of training data with known characteristic (Unsupervised).