Cancer classification is a topic of major interest in medicine since it allows accurate and efficient diagnosis and facilitates a successful outcome in medical treatments. Previous studies have classified human tumors using a large-scale RNA profiling and supervised Machine Learning (ML) algorithms to construct a molecular-based classification of carcinoma cells from breast, bladder, adenocarcinoma, colorectal, gastro esophagus, kidney, liver, lung, ovarian, pancreas, and prostate tumors. These datasets are collectively known as the 11_tumor database, although this database has been used in several works in the ML field, no comparative studies of different algorithms can be found in the literature. On the other hand, advances in both hardware and software technologies have fostered considerable improvements in the precision of solutions that use ML, such as Deep Learning (DL). In this study, we compare the most widely used algorithms in classical ML and DL to classify the tumors described in the 11_tumor database. We obtained tumor identification accuracies between 90.6% (Logistic Regression) and 94.43% (Convolutional Neural Networks) usingk-fold cross-validation. Also, we show how a tuning process may or may not significantly improve algorithms’ accuracies. Our results demonstrate an efficient and accurate classification method based on gene expression (microarray data) and ML/DL algorithms, which facilitates tumor type prediction in a multi-cancer-type scenario.

Advances in Deep Learning (DL) have provided alternative approaches to various complex problems, including the domain of spatial image steganalysis using Convolutional Neural Networks (CNN). Several CNN architectures have been developed in recent years, which have improved the detection accuracy of steganographic images. This work presents a novel CNN architecture which involves a preprocessing stage using filter banks to enhance steganographic noise, a feature extraction stage using depthwise and separable convolutional layers, and skip connections. Performance was evaluated using the BOSSbase 1.01 and BOWS 2 datasets with different experimental setups, including adaptive steganographic algorithms, namely WOW, S-UNIWARD, MiPOD, HILL and HUGO. Our results outperformed works published in the last few years in every experimental setting. This work improves classification accuracies on all algorithms and bits per pixel (bpp), reaching 80.3% on WOW with 0.2 bpp and 89.8% on WOW with 0.4 bpp, 73.6% and 87.1% on S-UNIWARD (0.2 and 0.4 bpp respectively), 68.3% and 81.4% on MiPOD (0.2 and 0.4 bpp), 68.5% and 81.9% on HILL (0.2 and 0.4 bpp), 74.6% and 84.5% on HUGO (0.2 and 0.4 bpp), using BOSSbase 1.01 test data.

Steganography consists of hiding messages inside some object known as a carrier in order to establish a covert communication channel so that the act of communication itself goes unnoticed by observers who have access to that channel. The steganalysis is dedicated to the detection of hidden messages using steganography; these messages can be implicit in different types of media, such as digital images, video files, audio files or plain text. Traditionally, steganalysis has been divided into two separate stages, the first stage consists of manual extraction of sophisticated features and the second stage is classification using Ensemble Classifiers or Support Vector Machines. In recent years, the development of Deep Learning has made it possible to unify and automate the two traditional stages into an end to end approach with promising results. This paper shows the evolution of steganalysis in recent years using the Deep Learning techniques. The results of these techniques have surpassed those obtained with conventional methods - Rich Models with Ensemble Classifiers - both in the spatial and frequency (JPEG) domains. Since 2014, researchers have used The Convolutional Neural Networks to solve this problem generating diverse architectures and strategies to improve the detection percentages of steganographic images on the last generation algorithms (WOW, S-UNIWARD, HUGO, J-UNIWARD, among others). The Deep Learning, being applied to steganalysis, is now in the process of construction and results so far are encouraging for researchers that are interested in the topic.

Artificial Intelligence (AI)-based systems and their increasingly common use have made it a ubiquitous technology; Machine Learning algorithms are present in streaming services, social networks, and in the health sector. However, implementing this emerging technology carries significant social and ethical risks and implications. Without ethical development of such systems, there is the potential for this technology to undermine people’s autonomy, privacy, and equity, even affecting human rights. Considering the approaches necessary for ethical development and effective governance of AI, such as ethical principles, guidelines, and technical tools, the question arises regarding the limitations of implementing these measures by the highly technical personnel involved in the process. In this context, we propose the creation of a typology that distinguishes the different stages of the AI life-cycle, the high-level ethical principles that should govern their implementation, and the tools with the potential to foster compliance with these principles, encompassing both technical and conceptual resources. In addition, this typology will include relevant information such as developmental level, related tasks, sectors, and language. Our research is based on a systematic review in which we identified 352 resources and tools. We expect this contribution to be valuable in promoting ethical AI development for developers and leaders who manage these initiatives. The complete typology and the comprehensive list of resources are available for consultation at https://ricardo-ob.github.io/tools4responsibleai.

Because of the promising results obtained by machine learning (ML) approaches in several fields, every day is more common, the utilization of ML to solve problems in bioinformatics. In genomics, a current issue is to detect and classify transposable elements (TEs) because of the tedious tasks involved in bioinformatics methods. Thus, ML was recently evaluated for TE datasets, demonstrating better results than bioinformatics applications. A crucial step for ML approaches is the selection of metrics that measure the realistic performance of algorithms. Each metric has specific characteristics and measures properties that may be different from the predicted results. Although the most commonly used way to compare measures is by using empirical analysis, a non-result-based methodology has been proposed, called measure invariance properties. These properties are calculated on the basis of whether a given measure changes its value under certain modifications in the confusion matrix, giving comparative parameters independent of the datasets. Measure invariance properties make metrics more or less informative, particularly on unbalanced, monomodal, or multimodal negative class datasets and for real or simulated datasets. Although several studies applied ML to detect and classify TEs, there are no works evaluating performance metrics in TE tasks. Here, we analyzed 26 different metrics utilized in binary, multiclass, and hierarchical classifications, through bibliographic sources, and their invariance properties. Then, we corroborated our findings utilizing freely available TE datasets and commonly used ML algorithms. Based on our analysis, the most suitable metrics for TE tasks must be stable, even using highly unbalanced datasets, multimodal negative class, and training datasets with errors or outliers. Based on these parameters, we conclude that the F1-score and the area under the precision-recall curve are the most informative metrics since they are calculated based on other metrics, providing insight into the development of an ML application.

One of the fields where Artificial Intelligence (AI) must continue to innovate is computer security. The integration of Wireless Sensor Networks (WSN) with the Internet of Things (IoT) creates ecosystems of attractive surfaces for security intrusions, being vulnerable to multiple and simultaneous attacks. This research evaluates the performance of supervised ML techniques for detecting intrusions based on network traffic captures. This work presents a new balanced dataset (IDSAI) with intrusions generated in attack environments in a real scenario. This new dataset has been provided in order to contrast model generalization from different datasets. The results show that for the detection of intruders, the best supervised algorithms are XGBoost, Gradient Boosting, Decision Tree, Random Forest, and Extra Trees, which can generate predictions when trained and predicted with ten specific intrusions (such as ARP spoofing, ICMP echo request Flood, TCP Null, and others), both of binary form (intrusion and non-intrusion) with up to 94% of accuracy, as multiclass form (ten different intrusions and non-intrusion) with up to 92% of accuracy. In contrast, up to 90% of accuracy is achieved for prediction on the Bot-IoT dataset using models trained with the IDSAI dataset.