Design Concept Generation With Variational Deep Embedding Over Comprehensive Optimization

This paper proposes a framework for generating design concepts through the loop of comprehensive exploitation and consequent exploration. The former is by any sophisticated optimization such as topology optimization with diversely different. The latter realization is due to the variational deep embedding (VaDE), a deep learning technique with classification capability. In the process of design concept generation first, exploitation through computational optimization generates various possibilities of design entities. Second, VaDE learns them. This learning encodes the clusters of similar entities over the latent space with smaller dimensions. The clustering result reveals some design concepts and identifies voids where as-yet-unrecognized design concepts are prospective. Third, the decoder of the learned VaDE generates some possibilities for new design entities. Forth such new entities are examined, and relevant new conditions will trigger further exploitation by the optimization. In this paper, this framework is implemented for and applied to the conceptual design problem of bridge structures. This application demonstrates that the framework can identify voids over the latent space and explore the possibility of new concepts. This paper brings up some discussion on the promises and possibilities of the proposed framework.

Hierarchical Self-supervised Augmented Knowledge Distillation

Knowledge distillation often involves how to define and transfer knowledge from teacher to student effectively. Although recent self-supervised contrastive knowledge achieves the best performance, forcing the network to learn such knowledge may damage the representation learning of the original class recognition task. We therefore adopt an alternative self-supervised augmented task to guide the network to learn the joint distribution of the original recognition task and self-supervised auxiliary task. It is demonstrated as a richer knowledge to improve the representation power without losing the normal classification capability. Moreover, it is incomplete that previous methods only transfer the probabilistic knowledge between the final layers. We propose to append several auxiliary classifiers to hierarchical intermediate feature maps to generate diverse self-supervised knowledge and perform the one-to-one transfer to teach the student network thoroughly. Our method significantly surpasses the previous SOTA SSKD with an average improvement of 2.56% on CIFAR-100 and an improvement of 0.77% on ImageNet across widely used network pairs. Codes are available at https://github.com/winycg/HSAKD.

Fine-tuning neural excitation/inhibition for tailored ketamine use in treatment-resistant depression

The glutamatergic modulator ketamine has been shown to rapidly reduce depressive symptoms in patients with treatment-resistant major depressive disorder (TRD). Although its mechanisms of action are not fully understood, changes in cortical excitation/inhibition (E/I) following ketamine administration are well documented in animal models and could represent a potential biomarker of treatment response. Here, we analyse neuromagnetic virtual electrode time series collected from the primary somatosensory cortex in 18 unmedicated patients with TRD and in an equal number of age-matched healthy controls during a somatosensory ‘airpuff’ stimulation task. These two groups were scanned as part of a clinical trial of ketamine efficacy under three conditions: (a) baseline; (b) 6–9 h following subanesthetic ketamine infusion; and (c) 6–9 h following placebo-saline infusion. We obtained estimates of E/I interaction strengths by using dynamic causal modelling (DCM) on the time series, thereby allowing us to pinpoint, under each scanning condition, where each subject’s dynamics lie within the Poincaré diagram—as defined in dynamical systems theory. We demonstrate that the Poincaré diagram offers classification capability for TRD patients, in that the further the patients’ coordinates were shifted (by virtue of ketamine) toward the stable (top-left) quadrant of the Poincaré diagram, the more their depressive symptoms improved. The same relationship was not observed by virtue of a placebo effect—thereby verifying the drug-specific nature of the results. We show that the shift in neural dynamics required for symptom improvement necessitates an increase in both excitatory and inhibitory coupling. We present accompanying MATLAB code made available in a public repository, thereby allowing for future studies to assess individually tailored treatments of TRD.

Constructing decision rules from naive bayes model for robust and low complexity classification

A large spectrum of classifiers has been described in the literature. One attractive classification technique is a Naïve Bayes (NB) which has been relayed on probability theory. NB has two major limitations: First, it requires to rescan the dataset and applying a set of equations each time to classify instances, which is an expensive step if a dataset is relatively large. Second, NB may remain challenging for non-statisticians to understand the deep work of a model. On the other hand, Rule-Based classifiers (RBCs) have used IF-THEN rules (henceforth, rule-set), which are more comprehensible and less complex for classification tasks. For elevating NB limitations, this paper presents a method for constructing a rule-set from the NB model, which serves as RBC. Experiments of the constructing rule-set have been conducted on (Iris, WBC, Vote) datasets. Coverage, Accuracy, M-Estimate, and Laplace are crucial evaluation metrics that have been projected to rule-set. In some datasets, the rule-set obtains significant accuracy results that reach 95.33 %, 95.17% for Iris and vote datasets, respectively. The constructed rule-set can mimic the classification capability of NB, provide a visual representation of the model, express rules infidelity with acceptable accuracy; an easier method to interpreting and adjusting from the original model. Hence, the rule-set will provide a comprehensible and lightweight model than NB itself.

A Complementary Sensory Tool for Children with Autism Spectrum Disorders

Background: Sensory integration (SI) issues are widely described in people with autism spectrum disorder (ASD), impacting in their daily life and occupations. To improve their quality of life and occupational performance, we need to improve clinical and educational evaluation and intervention processes. We aim to develop a tool for measuring SI issues for Spanish children and adolescents with ASD diagnosis, to be used as a complementary tool to complete the Rivière’s Autism Spectrum Inventory, a widely used instrument in Spanish speaking places to describe the severity of ASD symptoms, recently updated with a new sensory scale with three dimensions. Methods: 458 Spanish participants complemented the new questionnaire, initially formed by 73 items with a 1–5 Likert scale. Results: The instrument finally was composed of 41 items grouped in three factors: modulation disorders (13 items), discrimination disorders (13 items), and sensory-based motor disorders (15 items). The goodness-of-fit indices from factor analyses, reliability, and the analysis of the questionnaire’s classification capability offered good values. Conclusions: The new questionnaire shows good psychometric properties and seems to be a good complementary tool to complete new the sensory scale in the Rivière’s Autism Spectrum Inventory.

A novel facial image recognition method based on perceptual hash using quintet triple binary pattern

Image classification (categorization) can be considered as one of the most breathtaking domains of contemporary research. Indeed, people cannot hide their faces and related lineaments since it is highly needed for daily communications. Therefore, face recognition is extensively used in biometric applications for security and personnel attendance control. In this study, a novel face recognition method based on perceptual hash is presented. The proposed perceptual hash is utilized for preprocessing and feature extraction phases. Discrete Wavelet Transform (DWT) and a novel graph based binary pattern, called quintet triple binary pattern (QTBP), are used. Meanwhile, the K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) algorithms are employed for classification task. The proposed face recognition method is tested on five well-known face datasets: AT&T, Face94, CIE, AR and LFW. Our proposed method achieved 100.0% classification accuracy for the AT&T, Face94 and CIE datasets, 99.4% for AR dataset and 97.1% classification accuracy for the LFW dataset. The time cost of the proposed method is O(nlogn). The obtained results and comparisons distinctly indicate that our proposed has a very good classification capability with short execution time.