Explore the vast range of titles available.

Predictions obtained by, e.g., artificial neural networks have a high accuracy but humans often perceive the models as black boxes. Insights about the decision making are mostly opaque for humans. Particularly understanding the decision making in highly sensitive areas such as healthcare or finance, is of paramount importance. The decision-making behind the black boxes requires it to be more transparent, accountable, and understandable for humans. This survey paper provides essential definitions, an overview of the different principles and methodologies of explainable Supervised Machine Learning (SML). We conduct a state-of-the-art survey that reviews past and recent explainable SML approaches and classifies them according to the introduced definitions. Finally, we illustrate principles by means of an explanatory case study and discuss important future directions.

Understanding the properties of excited states of complex molecules is crucial for many chemical and physical processes. Calculating these properties is often significantly more resource-intensive than calculating their ground state counterparts. We present a quantum machine learning model that predicts excited-state properties from the molecular ground state for different geometric configurations. The model comprises a symmetry-invariant quantum neural network (NN) and a conventional NN and is able to provide accurate predictions with only a few training data points. The proposed procedure is fully NISQ compatible. This is achieved by using a quantum circuit that requires a number of parameters linearly proportional to the number of molecular orbitals, along with a parameterized measurement observable, thereby reducing the number of necessary measurements. We benchmark the algorithm on three different molecules with three different system sizes: H 2 with four orbitals, LiH with five orbitals, and H 4 with six orbitals. For these molecules, we predict the excited state transition energies and transition dipole moments. We show that, in many cases, the procedure is able to outperform various classical models (support vector machines, Gaussian processes, and NNs) that rely solely on classical features, by up to two orders of magnitude in the test mean squared error.

Objective was to analyze the role of PD-L1 and its relation to demographic, patho-clinical and outcome parameters in salivary gland carcinoma (SGC) patients. Patients treated for salivary gland carcinomas between 1994 and 2010 were included. A retrospective chart review for baseline characteristics, pathohistological, clinical and outcome data was performed. Immunohistochemistry for PD-L1 was performed using tissue microarrays. PD-L1 expression was assessed in tumor cells and tumor-infiltrating immune cells (TIIC) and statistical analysis with regard to baseline and outcome data was performed. Expression of PD-L1 (by means ≥1% of the cells with PD-L1 positivity) was present in the salivary gland carcinoma cells of 17%, in the TIIC of 20% and in both tumor cells and TIIC of 10% the patients. PD-L1 expression in tumor cells and both tumor cells and TIIC was related to tumor grading (p = 0.035 and p = 0.031, respectively). A trend towards higher grading was also seen for PD-L1 expression in TIICs (p = 0.058). Patients with salivary duct carcinomas and PD-L1 expressing TIICs showed a significantly worse DFS and OS (p = 0.022 and p = 0.003, respectively), those with both tumor cells and TIIC expressing PD-L1 a significantly worse DFS (p = 0.030). PD-L1 expression is present in 17% and 20% of salivary gland carcinoma cells and TIIC. Ten percent of the patient showed a PD-L1 positivity in both tumor cells and TIIC. This is related to high tumor grading and therefore might be a negative prognostic factor.

The advancing digitalisation is one of the great challenges of our times. Related activities also concern the wastewater sector. In the field of building construction, one emerging approach is building information modelling (BIM). The presented work investigates to which extent BIM practices have already found their way to wastewater management, and what kind of benefits and constraints are incorporated. Information is collected by means of a literature review and international expert surveys. Results indicate that several BIM-related key elements are already well established in the sector, but not necessarily in the intended manner. Consequently, the digital transition in the wastewater sector is not about replacing existing procedures and techniques but to rethink and optimise them. This primarily concerns data and information management in combination with the application of digital tools. Furthermore, wastewater management requires more integrated approaches, involving interdisciplinary/collaborative concepts and life cycle perspectives. Appropriate change management is necessary to give support and guidance to employees during the transition process. Furthermore, also from the political side, a clear definition and communication of the pursued digital vision is important. This article aims at stimulating discussion and research to optimise wastewater management from the digital perspective.

Manual inspection of workpieces in highly flexible production facilities with small lot sizes is costly and less reliable compared to automated inspection systems. Reinforcement Learning (RL) offers promising, intelligent solutions for robotic inspection and manufacturing tasks. This paper presents an RL-based approach to determine a high-quality set of sensor view poses for arbitrary workpieces based on their 3D computer-aided design (CAD). The framework extends available open-source libraries and provides an interface to the Robot Operating System (ROS) for deploying any supported robot and sensor. The integration into commonly used OpenAI Gym and Baselines leads to an expandable and comparable benchmark for RL algorithms. We give a comprehensive overview of related work in the field of view planning and RL. A comparison of different RL algorithms provides a proof of concept for the framework’s functionality in experimental scenarios. The obtained results exhibit a coverage ratio of up to 0.8 illustrating its potential impact and expandability. The project will be made publicly available along with this article.

Malignant melanoma, the deadliest form of skin cancer, is characterized by a predominant mutation in the BRAF gene 1 , 2 , 3 . Drugs that target tumours carrying this mutation have recently entered the clinic 4 , 5 , 6 , 7 . Accordingly, patients are routinely screened for mutations in this gene to determine whether they can benefit from this type of treatment. The current gold standard for mutation screening uses real-time polymerase chain reaction and sequencing methods 8 . Here we show that an assay based on microcantilever arrays can detect the mutation nanomechanically without amplification in total RNA samples isolated from melanoma cells. The assay is based on a BRAF -specific oligonucleotide probe. We detected mutant BRAF at a concentration of 500 pM in a 50-fold excess of the wild-type sequence. The method was able to distinguish melanoma cells carrying the mutation from wild-type cells using as little as 20 ng µl –1 of RNA material, without prior PCR amplification and use of labels. Microcantilever arrays are used to detect individual point mutations in a gene associated with melanoma cancer, offering a rapid test for deciding whether or not patients are eligible to receive drug treatment.

Machine learning (ML) has become a vital part in many aspects of our daily life. However, building well performing machine learning applications requires highly specialized data scientists and domain experts. Automated machine learning (AutoML) aims to reduce the demand for data scientists by enabling domain experts to build machine learning applications automatically without extensive knowledge of statistics and machine learning. This paper is a combination of a survey on current AutoML methods and a benchmark of popular AutoML frameworks on real data sets. Driven by the selected frameworks for evaluation, we summarize and review important AutoML techniques and methods concerning every step in building an ML pipeline. The selected AutoML frameworks are evaluated on 137 data sets from established AutoML benchmark suites.

Quantum machine learning (QML) models use encoding circuits to map data into a quantum Hilbert space. While it is well known that the architecture of these circuits significantly influences core properties of the resulting model, they are often chosen heuristically. In this work, we present a approach using reinforcement learning techniques to generate problem-specific encoding circuits to improve the performance of QML models. By specifically using a model-based reinforcement learning algorithm, we reduce the number of necessary circuit evaluations during the search, providing a sample-efficient framework. In contrast to previous search algorithms, our method uses a layered circuit structure that significantly reduces the search space. Additionally, our approach can account for multiple objectives such as solution quality and circuit depth. We benchmark our tailored circuits against various reference models, including models with problem-agnostic circuits and classical models. Our results highlight the effectiveness of problem-specific encoding circuits in enhancing QML model performance.

Oriented cell division is a fundamental mechanism to control asymmetric stem cell division, neural tube elongation and body axis extension, among other processes. During zebrafish gastrulation, when the body axis extends, dorsal epiblast cells display divisions that are robustly oriented along the animal-vegetal embryonic axis. Here, we use a combination of lipidomics, metabolic tracer analysis and quantitative image analysis to show that sphingolipids mediate spindle positioning during oriented division of epiblast cells. We identify the Wnt signaling as a regulator of sphingolipid synthesis that mediates the activity of serine palmitoyltransferase (SPT), the first and rate-limiting enzyme in sphingolipid production. Sphingolipids determine the palmitoylation state of the Anthrax receptor, which then positions the mitotic spindle of dividing epiblast cells. Our data show how Wnt signaling mediates sphingolipid-dependent oriented division and how sphingolipids determine Anthrax receptor palmitoylation, which ultimately controls the activation of Diaphanous to mediate spindle rotation and oriented mitosis. During development, oriented cell division is important to proper body axis extension. Here, the authors show that sphingolipids are required to direct spindle rotation and oriented mitosis via Anthrax receptor palmitoylation in zebrafish gastrulation.