Mining materials knowledge with concept lattice algorithm

In the sessions of Materials Genome Initiative (MGI) and Materials Informatics (MI) at recent international and domestic conferences, majority of research presentations are on material prediction using a combination of computational chemistry and machine learning. On the other hand, the contents incorporating high-throughput synthesis and evaluation experiments are only occasionally seen. It is also true that the number of samples that can be handled per hour is small compared to the former data science.However, if a series of experimental processes linked to synthesis, evaluation, and analysis is made user-friendly and high-throughput, it is possible to efficiently generate multi-condition and multi-component data in a unified experimental environment such as starting materials. As a result, even when creating a data set by extracting text from a paper, the data shortage can be supplemented with experimental data.Unlike thin films and polymers, powder synthesis is a heterogeneous reaction that requires the use of a wet process to perform high-throughput experiments. We have hitherto developed powder synthesis apparatus equipped with a micropump for solution dispensing, and high-throughput exploration system for powder and thin-film library based on electrostatic spray deposition consisted by combining multiple syringe pumps and a high-voltage power supply. By using these system, we have been exploring multi-component cathode materials for lithium ion battery and thermoelectric materials. In addition, recently, high-throughput experiments in a high-pressure environment of 200 MPa and 500ºC are possible.If the time required for physical property evaluation and data analysis is the same as before, even if only the synthesis based on experiments becomes high throughput, the attractiveness of MI will be weakened. Therefore, it is necessary to continue to develop jigs and software for efficient evaluation and analysis of huge libraries obtained in high-throughput experiments.In this presentation, we will introduce the exploration for A-site and B-site substitutes of perovskite-type CaMnO3, which is expected as an n-type thermoelectric material. In the preliminary search, it was confirmed that Ca1-x Bi x Mn1-y Ni y O3 contributes to the improvement of thermoelectric power by substituting elements at each site. As the next step, the amount of Bi substitution was determined to be within 10% from the combination of high-throughput synthesis and high-throughput Seebeck coefficient measurement. Synchrotron XRD and XAFS measurements were used to investigate the correlation between physical properties and crystal structure. We have succeeded in collecting data without filling the capillaries and making pellets by developing special tools. Synchrotron XRD and XAFS measurements were used to investigate the correlation between physical properties and crystal structure. We have succeeded in collecting data without filling the capillaries and making pellets by developing special tools. We also developed software that automates Rietveld analysis, although we need basic knowledge to set up an initial structural model. As a result, we have made it easier to generate data sets for visualizing crystallographic data and physical property data at once. In the obtained Bi-substituted CaMnO3 powder, when Bi was 8% or less, the conductivity and power factor increased due to the increase in carrier concentration. At Bi8% and higher, PF reaches its limit value, which involves a large change in the MnO6 octahedral plane bond distances Mn-O1 (1) and Mn-O1 (2), that is, a significant increase in octahedral distortion. High-throughput experiments including the collection of crystallographic information through the development of these technologies are expected to be an effective tool for future machine learning.

Chemical heuristics have been fundamental to the advancement of chemistry and materials science. These heuristics are typically established by scientists using knowledge and creativity to extract patterns from limited datasets. Machine learning offers opportunities to perfect this approach using computers and larger datasets. Here, we discuss the relationships between traditional heuristics and machine learning approaches. We show how traditional rules can be challenged by large-scale statistical assessment and how traditional concepts commonly used as features are feeding the machine learning techniques. We stress the waste involved in relearning chemical rules and the challenges in terms of data size requirements for purely data-driven approaches. Our view is that heuristic and machine learning approaches are at their best when they work together.

Concept lattice is an effective tool for data analysis and extracting classification rules. However, the classical concept lattice often produce a lot of redundant rules. Closed-label concept lattice realizes reduction of concept intention, which can be used to extract fewer rules than the classical concept lattice. This paper presents a method for classification rules extraction based on the closed-label concept lattice. Examples show that the proposed method is effective for extracting more concise classification rules.

Formal concept analysis based on fuzzy granularity base for different granulations

Innovative storage technology and the rising popularity of the Internet have generated an ever-growing amount of data. In this vast amount of data much valuable knowledge is available, yet it is hidden. The Support Vector Machine (SVM) is a state-of-the-art classification technique that generally provides accurate models, as it is able to capture non-linearities in the data. However, this strength is also its main weakness, as the generated non-linear models are typically regarded as incomprehensible black-box models. By extracting rules that mimic the black box as closely as possible, we can provide some insight into the logics of the SVM model. This explanation capability is of crucial importance in any domain where the model needs to be validated before being implemented, such as in credit scoring (loan default prediction) and medical diagnosis. If the SVM is regarded as the current state-of-the-art, SVM rule extraction can be the state-of-the-art of the (near) future. This chapter provides an overview of recently proposed SVM rule extraction techniques, complemented with the pedagogical Artificial Neural Network (ANN) rule extraction techniques which are also suitable for SVMs. Issues related to this topic are the different rule outputs and corresponding rule expressiveness; the focus on high dimensional data as SVM models typically perform well on such data; and the requirement that the extracted rules are in line with existing domain knowledge. These issues are explained and further illustrated with a credit scoring case, where we extract a Trepan tree and a RIPPER rule set from the generated SVM model. The benefit of decision tables in a rule extraction context is also demonstrated. Finally, some interesting alternatives for SVM rule extraction are listed.

Understanding the nature of chemical bonding and its variation in strength across physically tunable factors is important for the development of novel catalytic materials. One way to speed up this process is to employ machine learning (ML) algorithms with online data repositories curated from high-throughput experiments or quantum-chemical simulations. Despite the reasonable predictive performance of ML models for predicting reactivity properties of solid surfaces, the ever-growing complexity of modern algorithms, e.g., deep learning, makes them black boxes with little to no explanation. In this Perspective, we discuss recent advances of interpretable ML for opening up these black boxes from the standpoints of feature engineering, algorithm development, and post hoc analysis. We underline the pivotal role of interpretability as the foundation of next-generation ML algorithms and emerging AI platforms for driving discoveries across scientific disciplines.

The objective of this article is to investigate the problem of generating both positive and negative exact association rules when a formal context K of (positive) attributes is provided. A straightforward solution to this problem consists of conducting an apposition of the initial context K with its complementary context K, construct the concept lattice B(K|K) of apposed contexts and then extract rules. A more challenging problem consists of exploiting rules generated from each one of the contexts K and K to get the whole set of rules for the context K|K. In this paper, we analyze a set of identified situations based on distinct types of input, and come out with a set of properties. Obviously, the global set of (positive and negative) rules is a superset of purely positive rules (i.e., rules with positive attributes only) and purely negative ones since it generally contains mixed rules (i.e., rules in which at least a positive attribute and a negative attribute coexist). The paper presents also a set of inference rules to generate a subset of all mixed rules from positive, negative and mixed ones. Finally, two key conclusions can be drawn from our analysis: (i) the generic basis containing negative rules, ΣK, cannot be completely and directly inferred from the set ΣK of positive rules or from the concept lattice B(K), and (ii) the whole set of mixed rules may not be completely generated from ΣK alone, ΣK ∪ ΣK alone, or B(K) alone.

Autonomous on-line knowledge discovery and acquisition play an important role in cognitive informatics, cognitive computing, knowledge engineering, and computational intelligence. On the basis of the latest advances in cognitive informatics and denotational mathematics, this paper develops a web knowledge discovery engine for web document restructuring and comprehension, which decodes on-line knowledge represented in informal documents into cognitive knowledge represented by concept algebra and concept networks. A visualized concept network explorer and a semantic analyzer are implemented to capture and refine queries based on concept algebra. A graphical interface is built using concept and semantic models to refine users’ queries. To enable autonomous information restructuring by machines, a two-level knowledge base that mimics human lexical/syntactical and semantic cognition is introduced. The information restructuring model provides a foundation for automatic concept indexing and knowledge extraction from web documents. The web knowledge discovery engine extends machine learning capability from imperative and adaptive information processing to autonomous and cognitive knowledge processing with unstructured documents in natural languages.

Explainable artificial intelligence (XAI) is concerned with creating artificial intelligence that is intelligible and interpretable by humans. Many AI techniques build classifiers, some of which result in intelligible models, some of which don’t. Rule extraction from classifiers treated as black boxes is an important topic in XAI, that aims to find rule sets that describe classifiers and that are understandable to humans. Neural networks provide one type of classifier where it is difficult to explain why the inputs map to the decision; support vector machines provide a second example of this kind. A third type of classifier, k-nearest neighbour (k-NN), gives more interpretable classifiers, but suffers from performance problems as the model is little more than a representation of the training data. This work investigates a technique to extract rules from classifiers where the underlying problem’s feature space is Boolean, without looking at the inner structure of the classifier. For such a classifier with a small feature space, a Boolean function describing it can be directly calculated, whilst for a classifier with a larger feature space, a sampling method is investigated to produce rule-based approximations to the behaviour of the underlying classifier, with varying granularity, leading to XAI. The behaviour of the technique with neural network, support vector machine, and k-NN classifiers is experimentally assessed on a dataset of cross-site scripting (XSS) attacks, and proves to give very high accuracy and precision, often comparable to the classifier being approximated.KeywordsRule extractionExplainable AINeural networksXSS

This paper proposes a fuzzy rule extraction method from support vector machines. Support vector machines (SVM) are learning systems based on statistical learning theory that have been successfully applied to a wide variety of application. However, SVM are "black box" models, that is, they generate a solution with linear combination of kernel functions which has a quite difficult interpretation. Methods for rule extraction from trained SVM have already been proposed, however, the rules generated by these methods have, in their antecedents, intervals or functions. This format decreases the interpretability of the generated rules and jeopardizes the knowledge extraction capability. Hence, to increase the linguistic interpretability of the generated rules, we propose in this paper a methodology for extracting fuzzy rules from a trained SVM, where the rule's antecedents are associated with fuzzy sets.

Motivation: The recent development of high-throughput drug profiling (high content screening or HCS) provides a large amount of quantitative multidimensional data. Despite its potentials, it poses several challenges for academia and industry analysts alike. This is especially true for ranking the effectiveness of several drugs from many thousands of images directly. This paper introduces, for the first time, a new framework for automatically ordering the performance of drugs, called fractional adjusted bi-partitional score (FABS). This general strategy takes advantage of graph-based formulations and solutions and avoids many shortfalls of traditionally used methods in practice. We experimented with FABS framework by implementing it with a specific algorithm, a variant of normalized cut—normalized cut prime (FABS-NC′), producing a ranking of drugs. This algorithm is known to run in polynomial time and therefore can scale well in high-throughput applications.Results: We compare the performance of FABS-NC′ to other methods that could be used for drugs ranking. We devise two variants of the FABS algorithm: FABS-SVM that utilizes support vector machine (SVM) as black box, and FABS-Spectral that utilizes the eigenvector technique (spectral) as black box. We compare the performance of FABS-NC′ also to three other methods that have been previously considered: center ranking (Center), PCA ranking (PCA), and graph transition energy method (GTEM). The conclusion is encouraging: FABS-NC′ consistently outperforms all these five alternatives. FABS-SVM has the second best performance among these six methods, but is far behind FABS-NC′: In some cases FABS-NC′ produces over half correctly predicted ranking experiment trials than FABS-SVM.Availability: The system and data for the evaluation reported here will be made available upon request to the authors after this manuscript is accepted for publication.Contact: yxy128@berkeley.edu

Best practice in nursing: A concept analysis

With the rapid development of deep learning techniques, the applications have become increasingly widespread in various domains. However, traditional deep learning methods are often referred to as "black box" models with low interpretability of their results, posing challenges for their application in certain critical domains. In this study, we propose a comprehensive method for the interpretability analysis of sentiment models. The proposed method encompasses two main aspects: attention-based analysis and external knowledge integration. First, we train the model within sentiment classification and generation tasks to capture attention scores from multiple perspectives. This multi-angle approach reduces bias and provides a more comprehensive understanding of the underlying sentiment. Second, we incorporate an external knowledge base to improve evidence extraction. By leveraging character scores, we retrieve complete sentiment evidence phrases, addressing the challenge of incomplete evidence extraction in Chinese texts. Experimental results on a sentiment interpretability evaluation dataset demonstrate the effectiveness of our method. We observe a notable increase in accuracy by 1.3%, Macro-F1 by 13%, and MAP by 23%. Overall, our approach offers a robust solution for enhancing the interpretability of sentiment models by combining attention-based analysis and the integration of external knowledge.

Efficient concept lattice generation algorithm is a major problem for the applied research of formal concept analysis. Based on the basic property of concept, this paper put forwards and proves a theorem about the relationship of concepts, and then proposes an incremental concept lattice algorithm based on basic concepts. An example shows that the algorithm is feasible and effective.

Neural Network is a powerful pattern recognition algorithm capable of learning complex non-linear patterns. However, Neural Networks have a well-known drawback of being a “Black Box” learner that is not comprehensible or transferable thus making it unsuitable tasks that require a rational justification for making a decision. Rule Extraction methods can resolve this limitation by extracting comprehensible rules from a trained Network. In this paper, we present an algorithm called HERETIC that uses a symbolic learning algorithm (Decision Tree) on each unit of the Neural Network. Experiments and theoretical analysis show HERETIC generates highly accurate rules that closely approximates the Neural Network.