Search-based procedural content generation (PCG) is a well-known method for level generation in games. Its key advantage is that it is generic and able to satisfy functional constraints. However, due to the heavy computational costs to run these algorithms online, search-based PCG is rarely utilized for real-time generation. In this paper, we introduce mutation models, a new type of iterative level generator based on machine learning. We train a model to imitate the evolutionary process and use the trained model to generate levels. This trained model is able to modify noisy levels sequentially to create better levels without the need for a fitness function during inference. We evaluate our trained models on a 2D maze generation task. We compare several different versions of the method: training the models either at the end of evolution (normal evolution) or every 100 generations (assisted evolution) and using the model as a mutation function during evolution. Using the assisted evolution process, the final trained models are able to generate mazes with a success rate of 99% and high diversity of 86%. The trained model is many times faster than the evolutionary process it was trained on. This work opens the door to a new way of learning level generators guided by an evolutionary process, meaning automatic creation of generators with specifiable constraints and objectives that are fast enough for runtime deployment in games.

Driven by the goal of eradicating language barriers on a global scale, machine translation has solidified itself as a key focus of artificial intelligence research today. However, such efforts have coalesced around a small subset of languages, leaving behind the vast majority of mostly low-resource languages. What does it take to break the 200 language barrier while ensuring safe, high quality results, all while keeping ethical considerations in mind? In No Language Left Behind, we took on this challenge by first contextualizing the need for low-resource language translation support through exploratory interviews with native speakers. Then, we created datasets and models aimed at narrowing the performance gap between low and high-resource languages. More specifically, we developed a conditional compute model based on Sparsely Gated Mixture of Experts that is trained on data obtained with novel and effective data mining techniques tailored for low-resource languages. We propose multiple architectural and training improvements to counteract overfitting while training on thousands of tasks. Critically, we evaluated the performance of over 40,000 different translation directions using a human-translated benchmark, Flores-200, and combined human evaluation with a novel toxicity benchmark covering all languages in Flores-200 to assess translation safety. Our model achieves an improvement of 44% BLEU relative to the previous state-of-the-art, laying important groundwork towards realizing a universal translation system.

Artificial intelligence (AI) may be a distant and little-known subject for some people, but the reality is that it is much closer than many people believe. Through Artificial Intelligence, it is possible to combat violence against women; assist lawyers, law firms, and departments with document analysis and monitoring of changes in legislation; assist clients with financial matters; make farmers have better productivity; help the elderly to have a better quality of life, among thousands of other things. AI advancements will be no less significant. For example, AI will soon be able to accelerate drug discovery and green energy research. According to Andrew Ng, one of the world's leading AI experts, AI's advancement can add more than $10 trillion to the global economy by 2030. Many people fail to recognize that this isn't necessarily a bad thing or something to be afraid of.

Current methods to determine the energy efficiency of buildings require on-site visits of certified energy auditors which makes the process slow, costly, and geographically incomplete. To accelerate the identification of promising retrofit targets on a large scale, we propose to estimate building energy efficiency from widely available and remotely sensed data sources only, namely street view, aerial view, footprint, and satellite-borne land surface temperature (LST) data. After collecting data for almost 40,000 buildings in the United Kingdom, we combine these data sources by training multiple end-to-end deep learning models with the objective to classify buildings as energy efficient (EU rating A-D) or inefficient (EU rating E-G). After evaluating the trained models quantitatively as well as qualitatively, we extend our analysis by studying the predictive power of each data source in an ablation study. We find that the end-to-end deep learning model trained on all four data sources achieves a macro-averaged F1 score of 64.64% and outperforms the k-NN and SVM-based baseline models by 14.13 to 12.02 percentage points, respectively. Thus, this work shows the potential and complementary nature of remotely sensed data in predicting energy efficiency and opens up new opportunities for future work to integrate additional data sources.

In this work, we present the SOMOS dataset, the first large-scale mean opinion scores (MOS) dataset consisting of solely neural text-to-speech (TTS) samples. It can be employed to train automatic MOS prediction systems focused on the assessment of modern synthesizers, and can stimulate advancements in acoustic model evaluation. It consists of 20K synthetic utterances of the LJ Speech voice, a public domain speech dataset which is a common benchmark for building neural acoustic models and vocoders. Utterances are generated from 200 TTS systems including vanilla neural acoustic models as well as models which allow prosodic variations. An LPCNet vocoder is used for all systems, so that the samples' variation depends only on the acoustic models. The synthesized utterances provide balanced and adequate domain and length coverage. We collect MOS naturalness evaluations on 3 English Amazon Mechanical Turk locales and share practices leading to reliable crowdsourced annotations for this task. We provide baseline results of state-of-the-art MOS prediction models on the SOMOS dataset and show the limitations that such models face when assigned to evaluate TTS utterances.

SARS-COV-2 is a positive single-strand RNA-based macromolecule that has caused the death of more than 6.3 million people since June 2022. Moreover, by disturbing global supply chains through lockdown, the virus has indirectly caused devastating damage to the global economy. It is vital to design and develop drugs for this virus and its various variants. In this paper, we developed an in-silico study-based hybrid framework to repurpose existing therapeutic agents in finding drug-like bioactive molecules that would cure Covid-19. We employed the Lipinski rules on the retrieved molecules from the ChEMBL database and found 133 drug-likeness bioactive molecules against SARS coronavirus 3CL Protease. Based on standard IC50, the dataset was divided into three classes active, inactive, and intermediate. Our comparative analysis demonstrated that the proposed Extra Tree Regressor (ETR) based QSAR model has improved prediction results related to the bioactivity of chemical compounds as compared to Gradient Boosting, XGBoost, Support Vector, Decision Tree, and Random Forest based regressor models. ADMET analysis is carried out to identify thirteen bioactive molecules with ChEMBL IDs 187460, 190743, 222234, 222628, 222735, 222769, 222840, 222893, 225515, 358279, 363535, 365134 and 426898. These molecules are highly suitable drug candidates for SARS-COV-2 3CL Protease. In the next step, the efficacy of bioactive molecules is computed in terms of binding affinity using molecular docking and then shortlisted six bioactive molecules with ChEMBL IDs 187460, 222769, 225515, 358279, 363535, and 365134. These molecules can be suitable drug candidates for SARS-COV-2. It is anticipated that the pharmacologist/drug manufacturer would further investigate these six molecules to find suitable drug candidates for SARS-COV-2. They can adopt these promising compounds for their downstream drug development stages.

Learning from imbalanced data is a challenging task. Standard classification algorithms tend to perform poorly when trained on imbalanced data. Some special strategies need to be adopted, either by modifying the data distribution or by redesigning the underlying classification algorithm to achieve desirable performance. The prevalence of imbalance in real-world datasets has led to the creation of a multitude of strategies for the class imbalance issue. However, not all the strategies are useful or provide good performance in different imbalance scenarios. There are numerous approaches to dealing with imbalanced data, but the efficacy of such techniques or an experimental comparison among those techniques has not been conducted. In this study, we present a comprehensive analysis of 26 popular sampling techniques to understand their effectiveness in dealing with imbalanced data. Rigorous experiments have been conducted on 50 datasets with different degrees of imbalance to thoroughly investigate the performance of these techniques. A detailed discussion of the advantages and limitations of the techniques, as well as how to overcome such limitations, has been presented. We identify some critical factors that affect the sampling strategies and provide recommendations on how to choose an appropriate sampling technique for a particular application.

Information extraction methods proved to be effective at triple extraction from structured or unstructured data. The organization of such triples in the form of (head entity, relation, tail entity) is called the construction of Knowledge Graphs (KGs). Most of the current knowledge graphs are incomplete. In order to use KGs in downstream tasks, it is desirable to predict missing links in KGs. Different approaches have been recently proposed for representation learning of KGs by embedding both entities and relations into a low-dimensional vector space aiming to predict unknown triples based on previously visited triples. According to how the triples will be treated independently or dependently, we divided the task of knowledge graph completion into conventional and graph neural network representation learning and we discuss them in more detail. In conventional approaches, each triple will be processed independently and in GNN-based approaches, triples also consider their local neighborhood. View Full-Text

Principal component analysis (PCA), a ubiquitous dimensionality reduction technique in signal processing, searches for a projection matrix that minimizes the mean squared error between the reduced dataset and the original one. Since classical PCA is not tailored to address concerns related to fairness, its application to actual problems may lead to disparity in the reconstruction errors of different groups (e.g., men and women, whites and blacks, etc.), with potentially harmful consequences such as the introduction of bias towards sensitive groups. Although several fair versions of PCA have been proposed recently, there still remains a fundamental gap in the search for algorithms that are simple enough to be deployed in real systems. To address this, we propose a novel PCA algorithm which tackles fairness issues by means of a simple strategy comprising a one-dimensional search which exploits the closed-form solution of PCA. As attested by numerical experiments, the proposal can significantly improve fairness with a very small loss in the overall reconstruction error and without resorting to complex optimization schemes. Moreover, our findings are consistent in several real situations as well as in scenarios with both unbalanced and balanced datasets.

ImageNet-1k is a dataset often used for benchmarking machine learning (ML) models and evaluating tasks such as image recognition and object detection. Wild animals make up 27% of ImageNet-1k but, unlike classes representing people and objects, these data have not been closely scrutinized. In the current paper, we analyze the 13,450 images from 269 classes that represent wild animals in the ImageNet-1k validation set, with the participation of expert ecologists. We find that many of the classes are ill-defined or overlapping, and that 12% of the images are incorrectly labeled, with some classes having >90% of images incorrect. We also find that both the wildlife-related labels and images included in ImageNet-1k present significant geographical and cultural biases, as well as ambiguities such as artificial animals, multiple species in the same image, or the presence of humans. Our findings highlight serious issues with the extensive use of this dataset for evaluating ML systems, the use of such algorithms in wildlife-related tasks, and more broadly the ways in which ML datasets are commonly created and curated.