Procedural Dungeon Generation: A Survey

Procedural content generation (PCG) is a method of content creation entirely or partially done by computers. PCG is popularly employed in game development to produce game content, such as maps and levels. Representative examples of games using PCG are Rogue (1998), which introduced the rogue­like genre, and No Man’s Sky (2016), which generated whole worlds with fauna and flora. PCG may generate final contents, ready to be added to a game, or intermediate contents, which might be polished by human designers or work as an input level sketch to be interpreted by a level translator. In this paper, we survey the current state of procedural dungeon generation (PDG) research, a PCG subarea, applied in the context of games. For each work we selected in this survey, we examined and compared how they created game features, what type of level structure and representation they propose, which content generation strategy they applied, and, finally, we classify them according to the taxonomy of procedural content generation proposed by Togelius et al. (2016). The most relevant findings of our survey are: (1) PDG for 3D levels has been little explored; (2) few works supported levels with barriers, a game mechanic which temporarily blocks the player progression, and; (3) mixed-initiative approaches, i.e., software that helps human designers by making suggestions to the levels being created, are little explored.

New inverse DEA models for budgeting and planning

Data envelopment analysis (DEA) measures the efficiency score of a set of homogeneous decision-making units (DMUs) based on observed input and output. Considering input-oriented, the inverse DEA models find the required input level for producing a given amount of production in the current efficiency level. This article proposes a new form of the inverse DEA model considering income (for planning) and budget (for finance and budgeting) constraints. In contrast with the classical inverse model, both input and output levels are variable in proposed models to meet income (or budget) constraints. Proposed models help decision-makers (DMs) to find the required value of each input and each output's income share to meet the income or budget constraint. We apply the proposed model in the efficiency analysis of 58 supermarkets belonging to the same chain. However, these methods are general and can be used in the budgeting and planning process of any production system, including business sectors and firms that provide services.

Evaluating the performance of primary schools in India: evidence from West Bengal

PurposeIn the context of India, this article proposes an integrated multicriteria decision-making (MCDM) regression-based methodology to evaluate input-level performance of the schools and investigate the impact of this performance along with contextual factors, i.e. medium of instruction and location of the school, on the school's output level performance, i.e. student pass rate.Design/methodology/approachFirst, Shannon entropy-based approach is applied for the weight assignment to different parameters. Then, integrated VlseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR) technique for order preference by similarity to an ideal solution (TOPSIS)-based methodology is devised to measure the input-level performance of a school. Finally, multiple linear regression (MLR) analysis is incorporated to study the effect of input-level performance and above-mentioned contextual factors on the school's output-level performance.FindingsProposed methodology is applied to assess the input-level performance of 82,930 primary and secondary schools of West Bengal, India. All the factors have a significant impact on boys' pass rate, whereas only input-level performance and location of the school have a significant influence on the girls' pass rate.Practical implicationsThe entropy-based approach highlights the importance of scientific weight assignment. Integrated MCDM demonstrates the significance of aggregation due to the variation in scores related to input-level performance across the methods. Regression analysis facilitates the exploration of determinants influencing the output-level performance of the schools.Originality/valueThis work depicts a holistic picture of the performance measurement system of the schools. It encompasses scientific weight assignment to the evaluation criteria, integrated input-level performance assessment of the schools and investigation into the effect of this performance, as well as other contextual factors on the output level performance.

Learning Subject-Generalized Topographical EEG Embeddings Using Deep Variational Autoencoders and Domain-Adversarial Regularization

Two of the biggest challenges in building models for detecting emotions from electroencephalography (EEG) devices are the relatively small amount of labeled samples and the strong variability of signal feature distributions between different subjects. In this study, we propose a context-generalized model that tackles the data constraints and subject variability simultaneously using a deep neural network architecture optimized for normally distributed subject-independent feature embeddings. Variational autoencoders (VAEs) at the input level allow the lower feature layers of the model to be trained on both labeled and unlabeled samples, maximizing the use of the limited data resources. Meanwhile, variational regularization encourages the model to learn Gaussian-distributed feature embeddings, resulting in robustness to small dataset imbalances. Subject-adversarial regularization applied to the bi-lateral features further enforces subject-independence on the final feature embedding used for emotion classification. The results from subject-independent performance experiments on the SEED and DEAP EEG-emotion datasets show that our model generalizes better across subjects than other state-of-the-art feature embeddings when paired with deep learning classifiers. Furthermore, qualitative analysis of the embedding space reveals that our proposed subject-invariant bi-lateral variational domain adversarial neural network (BiVDANN) architecture may improve the subject-independent performance by discovering normally distributed features.

Handling missing MRI sequences in deep learning segmentation of brain metastases: a multicenter study

The purpose of this study was to assess the clinical value of a deep learning (DL) model for automatic detection and segmentation of brain metastases, in which a neural network is trained on four distinct MRI sequences using an input-level dropout layer, thus simulating the scenario of missing MRI sequences by training on the full set and all possible subsets of the input data. This retrospective, multicenter study, evaluated 165 patients with brain metastases. The proposed input-level dropout (ILD) model was trained on multisequence MRI from 100 patients and validated/tested on 10/55 patients, in which the test set was missing one of the four MRI sequences used for training. The segmentation results were compared with the performance of a state-of-the-art DeepLab V3 model. The MR sequences in the training set included pre-gadolinium and post-gadolinium (Gd) T1-weighted 3D fast spin echo, post-Gd T1-weighted inversion recovery (IR) prepped fast spoiled gradient echo, and 3D fluid attenuated inversion recovery (FLAIR), whereas the test set did not include the IR prepped image-series. The ground truth segmentations were established by experienced neuroradiologists. The results were evaluated using precision, recall, Intersection over union (IoU)-score and Dice score, and receiver operating characteristics (ROC) curve statistics, while the Wilcoxon rank sum test was used to compare the performance of the two neural networks. The area under the ROC curve (AUC), averaged across all test cases, was 0.989 ± 0.029 for the ILD-model and 0.989 ± 0.023 for the DeepLab V3 model (p = 0.62). The ILD-model showed a significantly higher Dice score (0.795 ± 0.104 vs. 0.774 ± 0.104, p = 0.017), and IoU-score (0.561 ± 0.225 vs. 0.492 ± 0.186, p < 0.001) compared to the DeepLab V3 model, and a significantly lower average false positive rate of 3.6/patient vs. 7.0/patient (p < 0.001) using a 10 mm3 lesion-size limit. The ILD-model, trained on all possible combinations of four MRI sequences, may facilitate accurate detection and segmentation of brain metastases on a multicenter basis, even when the test cohort is missing input MRI sequences.

A Family of High Voltage Gain Three-Level Step-Up Converters for Photovoltaic Module Integration Applications

This paper proposes a family of step-up three-level DC-DC converter topologies suitable for photovoltaic panel integration applications. The proposed family is suitable to convert the 10–30 V from photovoltaic panels to a 150 V direct current distribution bus. The proposed family enhances the three-level topology in terms of the voltage gain, power density, and filtering requirements at the input level. The filtration is reduced by interleaving. The three-level boost converter’s voltage gain is enhanced by utilizing several options such as switched capacitor cells, switched inductor cells, and flyback transformers or coupled inductors. The enhancement techniques are illustrated by providing the circuit diagram and a comparison of the voltage gain and the number of required components. An example converter of a hybrid three-level boost converter with a flyback transformer is presented to convert 20 V from a photovoltaic panel to a 400 V. The theory of operation and steady-state analysis are provided for the example converter operating in the continuous conduction mode. The converter is simulated to extract the power from three PVL-136 photovoltaic (PV) panels by applying a maximum power point tracking algorithm. The theory of operation and simulation are confirmed with an 80 W experimental prototype, which has an efficiency of around 95% at 40 W load power.