After three successful versions of the Workshop on Collaborative Technologies
and Data Science in Smart City Applications (CODASSCA 2018, 2020, and 2022) we
are glad to present proceedings of the fourth version held at the American
University of Armenia (AUA), October 3-6, 2024 in Yerevan, Armenia. Data
Science and Reliable Machine Learning (ML) presents seventeen selected and
carefully revised papers, which are available at the start of the workshop in
the Open Access Proceedings and as a printed version. Society, technologies,
and sciences are undergoing a rapid and revolutionary shift towards the
integration of artificial intelligence into every system that people use in
everyday life to create smart environments (SmE) through ambient intelligence
(AmI) in highly connected and collaborative scenarios. The main source and
asset for making smart systems is data, produced today in extraordinary large
quantities. Data volumes are growing rapidly due to environmental sensor
reading, sensor networks, broadband services, multimodal communication, whereas
pervasive and embedded computing is enhancing the capability of everyday
objects and easing collaboration among people. Data from all areas of daily
life that are increasingly accessible to a broad public enable the conception,
creation, calibration, and validation of a process or complex systems. However,
this requires international standards for data quality and access. Mobile
systems could enhance the possibilities available for designers and
practitioners. Effective analysis, quality assessment and utilization of big
data is a key factor for success in many business and service domains,
including the domain of smart systems. Major industrial domains are on the way
to perform this tectonic shift based on Big Data, Artificial Intelligence,
Collaborative Technologies, Smart Environments (SmE) supporting Virtual and
Mixed Reality Applications, Multimodal Interaction and Reliable Visual and
Cognitive Analytics. However, many requirements must be fulfilled and
complexities resolved before we can effectively and efficiently turn the huge
amount of generated data into information and knowledge. The first one is to
ensure data quality, which includes accuracy and integrity of the obtained
data, timely delivery, suitable quantity, etc. Privacy and security
requirements and thorough end-to-end rights complement realization and
deployment of modern design, software development and evaluation tools. The
second one is to create understandable models, which can turn data into
valuable information and then into knowledge. A general challenge is the low
interpretability of various artificial intelligence (AI) approaches and ML
models, for which it is important that data scientists design the models, users
understand results and developers debug and improve the models. The increasing
complexity, limited explainability, and interpretability of the complex ML
models make it difficult to address the emerging requirements for acceptance of
these models and hinders their applications in industrial and mission critical
scenarios. Therefore, explainability, interpretability, transparency, and
accountability of ML models and systems need to be further developed for an
effective use of AI technologies. They are a prerequisite for a reliable
application of AI within many problem areas, e.g., natural language processing,
risk prediction in healthcare, fault/anomaly detection, computer vision or
classification and regression under uncertainty, which are significant ML
tasks. Researchers and practitioners working on theoretical and practical
aspects of data science and reliable machine learning, as well as related and
fundamental topics of statistical analysis, information transfer and processing
were invited to attend this workshop. The volume consists of the front matter
and three thematic sections with seventeen peer-reviewed contributions. The
editors would like to express their gratitude to the Foundation for Armenian
Science and Technology, the German Research Foundation (DFG) and the German
Academic Exchange Service (DAAD) for funding their activities; to Yanling Chen,
Rubina Danilova, Amalya Hambardzumyan, Ashot Harutyanyan, and Gregor Schiele
for their ongoing encouragement and support, our reviewers and to all
participants for their presentations and contributions to the workshop and this
proceedings volume. We are particularly grateful to the director of the
Martiros Sarian HouseMuseum in Yerevan, Mrs. Rouzan Sarian, for allowing us to
use her grandfather's painting of Aragats as the cover picture for this volume.
Yerevan, Tsukuba, Duisburg, October 2024