Machine learning and deep learning

Abstract

Today, intelligent systems that offer artificial intelligence capabilities often rely on machine learning. Machine learning describes the capacity of systems to learn from problem-specific training data to automate the process of analytical model building and solve associated tasks. Deep learning is a machine learning concept based on artificial neural networks. For many applications, deep learning models outperform shallow machine learning models and traditional data analysis approaches. In this article, we summarize the fundamentals of machine learning and deep learning to generate a broader understanding of the methodical underpinning of current intelligent systems. In particular, we provide a conceptual distinction between relevant terms and concepts, explain the process of automated analytical model building through machine learning and deep learning, and discuss the challenges that arise when implementing such intelligent systems in the field of electronic markets and networked business. These naturally go beyond technological aspects and highlight issues in human-machine interaction and artificial intelligence servitization.

A survey of image labelling for computer vision applications

“…In consequence, an industry of service providers has emerged already and will further proliferate, ranging from simple labelling platforms through outsourcing services to so-called "full-stack AI"-platforms offering the benefits of transfer learning (Janiesch et al, 2021), comprising a vast and diverse landscape for further research.…”

“…Inspired by information processing in biological systems, they consist of multiple interconnected processing units that forward signals using weights and activation functions. Artificial neural networks learn by processing many examples and iteratively adjusting the internal weights according to the difference to the known outcomes (Janiesch et al, 2021).…”

“…To realise such scenarios, modern CV systems rely on advanced methods from the field of machine learning (ML). Thus, instead of manually defining rules and patterns to execute vision-based tasks, ML models are able to process spatial information in raw image data and learn patterns automatically that are relevant for prediction tasks like recognising, localising, and segmenting visual objects (Janiesch et al, 2021).…”

A Survey on Recent Named Entity Recognition and Relationship Extraction Techniques on Clinical Texts

“…ANNs can be very deep, depending on the number of hidden layers between the input and the output, leading to deep learningbased methods. The differences between the traditional shallow methods and ANNs are surveyed by Janiesch et al (2021) [89]. Examples of traditional algorithms include but are not limited to Support Vector Machines (SVM), Linear Regression, Logistic Regression, Naive Bayes, Linear Discriminant Analysis, Decision Trees, K-Nearest Neighbor (KNN), Node2vec, etc., whereas Dense Neural Networks (DNN), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Graph Neural Networks (GNN), autoencoders, etc.…”

Electronic Markets on digital platforms and AI

“…In this vein, industry-specific cloud solutions were launched by platform providers, for example, Microsoft's Cloud for Healthcare or Amazon's initiatives for healthcare and life sciences (Healthlake), finance (Finspace) or manufacturing companies (Smart Factory) (Sawers, 2021). They point in the direction of AI-as-a-service offerings (Janiesch et al, 2021). Irrespective if these AI functionalities are applied internally or placed as services on the (external) market, a key question refers to traceability and the transparency of their behavior.…”