Search

The optimisation of a machine learning (ML) solution is a core research problem in the field of automated machine learning (AutoML). This process can require searching through complex configuration spaces of not only ML components and their hyperparameters but also ways of composing them together, i.e. forming ML pipelines. Optimisation efficiency and the model accuracy attainable for a fixed time budget suffer if this pipeline configuration space is excessively large. A key research question is whether it is both possible and practical to preemptively avoid costly evaluations of poorly performing ML pipelines by leveraging their historical performance for various ML tasks, i.e. meta-knowledge. This paper approaches the research question by first formulating the problem of configuration space reduction in the context of AutoML. Given a pool of available ML components, it then investigates whether previous experience can recommend the most promising subset to use as a configuration space when initiating a pipeline composition/optimisation process for a new ML problem, i.e. running AutoML on a new dataset. Specifically, we conduct experiments to explore (1) what size the reduced search space should be and (2) which strategy to use when recommending the most promising subset. The previous experience comes in the form of classifier/regressor accuracy rankings derived from either (1) a substantial but non-exhaustive number of pipeline evaluations made during historical AutoML runs, i.e. 'opportunistic' meta-knowledge, or (2) comprehensive cross-validated evaluations of classifiers/regressors with default hyperparameters, i.e. 'systematic' meta-knowledge. Overall, numerous experiments with the AutoWeka4MCPS package, including ones leveraging similarities between datasets via the relative landmarking method, suggest that (1) opportunistic/systematic meta-knowledge can improve ML outcomes, typically in line with how relevant that meta-knowledge is, and (2) configuration-space culling is optimal when it is neither too conservative nor too radical. However, the utility and impact of meta-knowledge depend critically on numerous facets of its generation and exploitation, warranting extensive analysis; these are often overlooked/underappreciated within AutoML and meta-learning literature. In particular, we observe strong sensitivity to the 'challenge' of a dataset, i.e. whether specificity in choosing a predictor leads to significantly better performance. Ultimately, identifying 'difficult' datasets, thus defined, is crucial to both generating informative meta-knowledge bases and understanding optimal search-space reduction strategies. • Exploring characteristics of ML problems and how these impact reduction strategies. • Assessing the advantages and disadvantages of constructing meta-knowledge. • Investigating the underlying mechanism of search space reduction strategies. [ABSTRACT FROM AUTHOR]

We report a rare case of primary malignant melanoma with inguinal lymph node metastasis in the male urethra. A 57-year-old male patient presented with a small tumor on the ventral surface of the penis, which was discovered 5 months ago and did not cause pain or discomfort. In the past month, the patient has developed symptoms of urinary incontinence. MRI and PET/CT scans revealed a primary tumor in the penile urethra, but no metastases were found. The patient underwent partial penectomy surgery and laparoscopic bilateral inguinal lymphadenectomy within one month. The pathological combined with immunohistochemical staining confirmed primary malignant melanoma in the urethra with right inguinal lymph node metastasis. Despite complying with surgical and immunotherapy treatment with Pembrolizumab for 18 cycles, the patient was diagnosed with recurrent cancer in the penile stump after 05 months and he passed away after 18 months. Urethral melanoma is a rare and highly invasive type of cancer. It was often diagnosed at a late stage because the initial symptoms were not obvious in the lower urinary tract. Additionally, cancer progressed very quickly, making it difficult to treat. Urethral melanoma, if detected at a late stage with lymph node metastasis, has a significantly poor prognosis irrespective of the treatment method employed. However, to our best knowledge, very few publications can be found on this disease, and the strategic treatment remained unknown. • The initial PET/CT scan did not reveal any metastases in the lymph nodes. However, after undergoing endoscopic inguinal lymphadenectomy surgery, we found lymph node metastasis. • Despite strict and aggressive adherence to surgery (partial penectomy and endoscopic inguinal lymphadenectomy) and immunotherapy with Pembrolizumab, the disease seemed intractable and recurred after 5 months, death after 18 months. • Case report of primary malignant melanoma of the male urethra, the first of its kind in Vietnam in English literature. [ABSTRACT FROM AUTHOR]

• Numerical model was validated from the full-scale experimental tests of RE structures. • Effect of semi-circular, parabolic and rectangular opening was analyzed and compared. • Arch-type openings is more stable and improves 30 % carrying capacity compared to rectangular forms. • A timber lintel for a rectangular opening increases the ultimate load by 102 % Rammed earth (RE) is a centuries-old construction technique that uses renewable resources found in the surrounding environment. As for concrete or timber houses, doors and windows are mandatory. Until date, the effect of openings on the mechanical behaviour of RE structures has received little attention. The primary contribution of this research is to assess the impact of opening on the behaviour of RE structures. A three-dimensional finite element models of RE walls were developed and calibrated against three experimental tests by Lilley and Robinson (1995). In the experimental study, RE walls with the geometrical dimensions 2285 mm × 1220 mm × 305 mm (length × height × thickness) were constructed from unstabilized soil and evaluated under in-plane compression loading. The considered walls have two types of openings: arch-type and rectangular shape. The numerical results obtained through non-linear finite element method analysis demonstrate the pertinence of the proposed model for quantifying bearing capacity and failure modes. Using the verified numerical model, a parametric study was carried out to investigate the effects of material characteristics, the geometrical characteristics of the opening and finally the improvement options using lintel reinforcements. [ABSTRACT FROM AUTHOR]

The aim of this paper is to study the density of solutions to stochastic Volterra integro-differential equations with multifractional noises. We prove the existence of the density of solutions and show that the density is bounded by lower and upper Gaussian bounds. [ABSTRACT FROM AUTHOR]

Abstract: In this paper we consider a stochastic Ginzburg–Landau equation with impulsive effects. We first prove the existence and uniqueness of the global solution which can be explicitly represented via the solution of a stochastic equation without impulses. Then, based on our obtained result, we study the qualitative properties of the solution, including the boundedness of moments, almost surely exponential convergence and pathwise estimations. Finally, we give a first attempt to study a fractional version of impulsive stochastic Ginzburg–Landau equations. [Copyright &y& Elsevier]

Automated machine learning pipeline (ML) composition and optimisation aim at automating the process of finding the most promising ML pipelines within allocated resources (i.e., time, CPU and memory). Existing methods, such as Bayesian-based and genetic-based optimisation, which are implemented in Auto-Weka, Auto-sklearn and TPOT, evaluate pipelines by executing them. Therefore, the pipeline composition and optimisation of these methods frequently require a tremendous amount of time that prevents them from exploring complex pipelines to find better predictive models. To further explore this research challenge, we have conducted experiments showing that many of the generated pipelines are invalid in the first place, and attempting to execute them is a waste of time and resources. To address this issue, we propose a novel method to evaluate the validity of ML pipelines, without their execution, using a surrogate model (AVATAR). The AVATAR generates a knowledge base by automatically learning the capabilities and effects of ML algorithms on datasets' characteristics. This knowledge base is used for a simplified mapping from an original ML pipeline to a surrogate model which is a Petri net based pipeline. Instead of executing the original ML pipeline to evaluate its validity, the AVATAR evaluates its surrogate model constructed by capabilities and effects of the ML pipeline components and input/output simplified mappings. Evaluating this surrogate model is less resource-intensive than the execution of the original pipeline. As a result, the AVATAR enables the pipeline composition and optimisation methods to evaluate more pipelines by quickly rejecting invalid pipelines. We integrate the AVATAR into the sequential model-based algorithm configuration (SMAC). Our experiments show that when SMAC employs AVATAR, it finds better solutions than on its own. This is down to the fact that the AVATAR can evaluate more pipelines within the same time budget and allocated resources. • We propose the AVATAR to evaluate the pipeline validity using a surrogate model. • Experiments show that invalid pipelines may result in overall bad performance. • Experiments show the efficiency of multiple configuration initialisation of SMAC. [ABSTRACT FROM AUTHOR]

The paper, using a three-wave unbalanced panel of 3252 observations of young small and medium-sized firms in 2011–2013, examines the effect of direct linkages between firms with foreign direct investment and young small and medium-sized firms on technology adoption strategies and the further influence of technology transfers from such linkages on technology adoption strategies. Moreover, the paper analyzes the extent that economic obstacles may cause young small and medium-sized firms to choose different adoptions. Our analysis shows that exporting firms do not tend to conduct embodied backward/forward adoptions, but more likely adopt the embodied backward purchasing. In addition, the impact of competitiveness follows an inverse U-shaped pattern for the embodied backward adoption, but a U-shaped pattern for the disembodied adoption. In terms of market power, there exists an inversed U-shaped pattern for the embodied backward adoption. Under the impact of foreign direct investment (FDI) linkages and vertical spillovers, it is found that technology transfer through backward/forward linkages is associated with the embodied ones, whereas a linkage with FDI domestic customers/suppliers is less likely associated with the embodied ones. In addition, under technology transfer, firms facing economic constraints may overcome these by looking for other financial sources and embodied technology transfer. The paper suggests the path for FDI firms, young small and medium-sized enterprises (SMEs), and technology adoption strategies in the future. [ABSTRACT FROM AUTHOR]

This study presents the sulfate resistance of various low energy no-cement mortars made with SFC binder which was fabricated by purely combining three industrial by-products of ground granulated blast furnace slag (GGBFS) (S), Class F fly ash (FFA) (F), and circulating fluidized bed combustion fly ash (CFA) (C). The sulfate resistance of the mortar was determined based on the effects of sodium sulfate solution (NaSO 4 , 5%) on its expansion, strength loss, and phase changes. Experimental results showed that the SFC mortars had much better sulfate resistance than that of ordinary Portland cement (OPC) mortars. The increase of FFA amount significantly increased the sulfate resistance of SFC cement mortars with lower expansion and relatively high compressive strengths. The diffusing ions such as Na + and SO 4 2− from the sulfate environment noticeably activated the pozzolanic reaction of FFA in the hardened SFC binder pastes resulting in stronger microstructure of the mortar specimens. On the contrary, these ions adversely caused undesired expansive damage of OPC mortars resulting from the formulation of excessive ettringite (AFt) and gypsum precipitation. [ABSTRACT FROM AUTHOR]