The Internet of things promises a continuous flow of data where traditional database and data-mining methods cannot be applied. This paper presents improvements on the Ubiquitous Self-Organized Map (UbiSOM), a novel variant of the well-known Self-Organized Map (SOM), tailored for streaming environments. This approach allows ambient intelligence solutions using multidimensional clustering over a continuous data stream to provide continuous exploratory data analysis. The average quantization error and average neuron utility over time are proposed and used to estimating the learning parameters, allowing the model to retain an indefinite plasticity and to cope with changes within a multidimensional data stream. We perform parameter sensitivity analysis and our experiments show that UbiSOM outperforms existing proposals in continuously modeling possibly non-stationary data streams, converging faster to stable models when the underlying distribution is stationary and reacting accordingly to the nature of the change in continuous real world data streams.
Competition is evident throughout our society, lives and work. “It transcends time and place, as well as all manners of people”. In this paper, we present a study and a framework proposal for a mixed collaborative-competitive learning environment, applied to a programming course in a University level engineering program. While neither collaboration, nor competition, is inherently good or bad in supporting the learning process, the way teachers employ these strategies in classrooms determines their value in preparing “soon-to-be” professionals. We believe this mixed-based learning approach best serves students as they are able to achieve academic success both in working with others and on an individual basis. Besides the traditional theoretical lessons and laboratory practices of the course, the integration of the framework automates all the support and evaluation processes. It implements both an individual competitive setting and a collaborative one, where intergroup competition occurs, by means of two interconnected modules. We discuss our results and findings in its preliminary use, concluding that the framework promises to have a good impact in the way we teach some programming courses in our Institute.
Data mining models are frequently used to represent and summarize meaningful properties in data. However such models are usually not suitable for interactive data exploration and visualization. This paper proposes the use of multidimensional projection together with the Ubiquitous Self-Organizing Map algorithm (UbiSOM), a novel variant of the well-known self-organizing map algorithm that was specifically tuned for stream data-mining. The resulting high-dimensional projection system is then studied for interactive data analysis and visualization. A prototype was developed where, at each moment, the user can visualize the information from different perspectives. Direct interaction with the system during stream processing is possible both by changing the projection, by optimizing the projection view for maximizing variance or by filtering the incoming data series. Experiments in two distinct datasets show the importance and relevance of conjoining multidimensional data projection with UbiSOM.
Blended learning refers to a mixing of different learning environments. It combines traditional classroom learning with modern computer-mediated activities. The present work presents the implementation of a Web-based system to help the application of blended learning to enhance learning in math courses of engineering degrees. The system is being integrated in the Moodle platform, which is used as the aiding electronic teaching system at our Polytechnic University. Our focus is on helping math learning, but we can apply it to courses that use similar platform approaches. The main goal of this paper is to present the system, focusing on the tool that facilitates the implementation of assessment moments, and discuss preliminary results according to teachers.
As the volume and complexity of data streams continue to increase, exploratory cluster analysis is becoming increasingly important. In this chapter, the authors explore the use of artificial neural networks (ANNs), particularly self-organizing maps (SOMs), for this purpose. They propose additional methodologies, including concept drift detection, as well as distributed and collaborative learning strategies and introduce a new open-source Java ANN library, designed to support practical applications of SOMs across various domains. By following our tutorial, users will gain practical insights into visualizing and analyzing these challenging datasets, enabling them to harness the full potential of our approach in their own projects. Overall, this chapter aims to provide readers with a comprehensive understanding of SOMs and their place within the broader context of artificial neural networks. Furthermore, we offer practical guidance on the effective development and utilization of these models in real-world applications. Request access from your librarian to read this chapter's full text.