Efficient management of cloud storage resources requires intelligent tier allocation strategies that balance cost optimization with performance requirements. While previous approaches have focused on binary classification schemes for storage tiering, real-world scenarios demand more granular solutions that can adapt to diverse user preferences and workload characteristics. This paper extends our previous work on access frequency prediction by proposing a comprehensive multiclass machine learning framework for intelligent cloud storage tiering. The proposed framework incorporates a novel three-tier classification system ( Cold / Warm / Hot ) and integrates user-centric preferences through a cost-weight parameter, enabling dynamic adaptation to varying preferences along the cost-latency spectrum. We demonstrate the framework’s effectiveness through extensive experiments on real-world access patterns, where we assess the performance of thirteen machine learning algorithms under various user preference profiles. The results show that our multiclass approach achieves cost reductions of up to 40% compared to a static tiering strategy, while providing Pareto-optimal solutions for different user profiles. Through comprehensive Pareto frontier analysis, we demonstrate the framework’s ability to provide transparent trade-off visualization, enabling informed decision-making for cloud storage administrators. Our main contributions are: a multiclass classification approach for storage tiering, the integration of user preferences via parameterized optimization, a comparative analysis of multiple algorithms across different preference configurations, and a practical validation of the framework’s applicability in production cloud storage environments.

Cloud storage providers typically offer multiple tiers with differing performance and cost. Classifying data into correct tiers is challenging, given evolving access patterns. This paper presents a supervised learning framework to predict object access frequency, thus enabling cost-effective tier allocations. Using real-world Dropbox traces, our experiments show up to 37% cost savings compared to an online tiering baseline. We evaluate multiple machine learning methods and time-window strategies, demonstrating the trade-offs between cost optimization and recall (to avoid misclassifving frequently accessed objects).

One of the main events involving the world economy in 2022 was the beginning of the war between Russia and Ukraine. This event offers an opportunity to analyze how a large-magnitude world event can affect the use of cryptocurrencies. Ethereum is one of the most prominent and widely used cryptocurrency platforms and, as such, provides a valuable case study for this scenario. This work investigates the behavior of accounts and their transactions on the Ethereum network during this event. For this purpose, we collect all Ethereum transactions during two distinct periods: (i) during the month the conflict began, and (ii) during the previous year. We organized a dataset with the accounts involved in these transactions and the subset of these accounts that interacted with a service within Ethereum named Flashbots Auction. Flashbots Auction is crucial as it addresses issues regarding transaction ordering and miners exploiting that ordering to make profit. Then, we model temporal graphs in which each vertex represents an account, and each edge represents a transaction between two accounts. We analyzed the behavior of these accounts via graph metrics for both groups during each observed time window. The results show changes in account behavior and activity, as well as variations in daily transaction volume.

This work presents a method for volumetric seismic data compression by coupling a 3-D convolution-based autoencoder to a generative adversarial network (GAN). The main challenge of 3-D convolutional autoencoders for data compression is how to fully exploit volumetric redundancy while keeping reasonable latent representation dimensions. Our method is based on a convolutional neural network for seismic data compression called 3DSC. Its encoder and decoder use 3-D convolutions and are connected by a latent representation with the same dimensions as its 2-D network counterparts. Our main hypothesis is that the 3DSC architecture can be improved by adversarial training. We, thus, propose a new 3-D-based seismic data compression method (3DSC-GAN) by coupling the 3DSC network to a GAN. The seismic data decoder is used as a generator of poststack data that are integrated with a discriminator module to better exploit 3-D redundancy. Results show that our method outperforms previous seismic data compression methods for very low target bit rates, increasing the peak signal-to-noise ratio (PSNR) with fairly high visual quality.

Recent deep learning techniques have achieved satisfactory results for various image-related problems. However, many research questions remain open in tasks involving video sequences. Several applications demand the understanding of complex events in videos, such as traffic monitoring, person re-identification, security and surveillance. In this work, we address the problem of human action recognition in videos through a multi-stream network that incorporates both spatial and temporal information. The main contribution of our work is a stream based on a new variant of the visual rhythm, called Learnable Visual Rhythm (LVR). We employ a deep network to extract features from the video frames in order to generate the rhythm. The features are collected at multiple depths of the network to enable the analysis of different abstraction levels. This strategy significantly outperforms the handcrafted version on the UCF101 and HMDB51 datasets. Experiments conducted on these datasets show that our final multi-stream network achieved competitive results compared to state-of-the-art approaches.

The concern about security and integrity of confidential messages has always been target of study. One of the most used techniques for this purpose is the encryption, which consists of hiding a message, making it incomprehensible, ensuring that only the sender and the addressee have access to its original content. In this process, the algorithm applies a key on the message, which aims to hide the message to send it or reveal it when it arrives at destination. In this paper we present a solution on this domain based on the use of metaheuristics for the design of unidimensional chaotic cellular automata used as a key generator for stream cipher algorithms.

The Bayes-optimal classifier is defined as a classifier that induces an hypothesis able to minimize the prediction error for any given sample in binary classification problems. Finding the Bayes-optimal classifier is an intractable problem. It is known that it is approximately equivalent to the center of mass of the version space, which is given by the set of all classifiers consistent with the training set. Previously solutions to find the center of mass are not feasible, as they present a high computational cost, and do not work properly in non-linear separable problems. Aiming to solve these problems, this paper presents the Dual Version Space Reduction Machine (Dual VSRM), an effective kernel method to approximate the center of mass of the version space. The Dual VSRM algorithm employs successive reductions of the version space based on an oracle's decision. As an oracle, we propose the Ensemble of Dissimilar Balanced Kernel Perceptrons (EBPK). EBPK enhances the accuracy of each individual classifier by balancing the final hyperplane solution while maximizing the diversity of its components by applying a dissimilarity measure. In order to evaluate the proposed methods, we conduct an experimental evaluation on 7 datasets. We compare the performance of our proposed methods against several baselines. Our results for EBKP indicate the strategies for improving individual accuracy and diversity of the ensemble components work properly. Also, the Dual VSRM consistently outperforms the baselines, indicating that the proposed method generates a better approximation to the center of mass.