While some studies have proven that Swin Transformer (Swin) with window self-attention (WSA) is suitable for single image super-resolution (SR), the plain WSA ignores the broad regions when reconstructing high-resolution images due to a limited receptive field. In addition, many deep learning SR methods suffer from intensive computations. To address these problems, we introduce the N-Gram context to the low-level vision with Transformers for the first time. We define N-Gram as neighboring local windows in Swin, which differs from text analysis that views N-Gram as consecutive characters or words. N-Grams interact with each other by sliding-WSA, expanding the regions seen to restore degraded pixels. Using the N-Gram context, we propose NGswin, an efficient SR network with SCDP bottleneck taking multi-scale outputs of the hierarchical encoder. Experimental results show that NGswin achieves competitive performance while maintaining an efficient structure when compared with previous leading methods. Moreover, we also improve other Swin-based SR methods with the N-Gram context, thereby building an enhanced model: SwinIR-NG. Our improved SwinIR-NG out-performs the current best lightweight SR approaches and establishes state-of-the-art results. Codes are available at https://github.com/rami0205/NGramSwin.

As multimedia content often contains noise from intrinsic defects of digital devices, image denoising is an important step for high-level vision recognition tasks. Although several studies have developed the denoising field employing advanced Transformers, these networks are too momory-intensive for real-world applications. Additionally, there is a lack of research on lightweight denosing (LWDN) with Transformers. To handle this, we provide seven comparative baseline Transformers for LWDN, serving as a foundation for future research. We also demonstrate the parts of randomly cropped patches significantly affect the denoising performances during training. While previous studies have overlooked this aspect, we aim to train our baseline Transformers in a truly fair manner. Furthermore, we conduct empirical analyses of various components to determine the key considerations for constructing LWDN Transformers. Codes are available at https://github.com/rami0205/LWDN.

We propose a novel stock movement prediction model that learns Multi-Time Contexts and Randomness (MTC-R). We assume that the stock price movements are affected by (1) combination of time-based momentums and (2) tradings by noise traders which cause randomness in a stock market. MTC-R has three main procedures for modeling our hypothesis. First, the model learns time representations using time2vec and encodes the multi-time views using a GRU network. Second, MTC-R generates the multi-time contexts using a multi-head attention mechanism and inserts the randomness. Third, a loss function is designed to learn temporal differences between the results of the second and the current time-embedded vector. Our model improves the prediction results in terms of accuracy and the Matthews correlation coefficient on six benchmark datasets compared to baseline models. Furthermore, MCT-R shows the effectiveness of the prediction results using cumulative returns in portfolio trading simulations.

Graph-based semi-supervised learning has recently come into focus for to its two defining phases: graph construction, which converts the data into a graph, and label inference, which predicts the appropriate labels for unlabeled data using the constructed graph. And the label inference is based on the smoothness assumption of semi-supervised learning. In this study, we propose an enhanced label inference approach which incorporates the importance of each vertex into the existing inference algorithms to improve the prediction capabilities of the algorithms. We also present extensions of three algorithms which are capable of taking the vertex importance variable to apply in learning. Experiments show that our algorithms perform better than the base algorithms on a variety of datasets, especially when the data is less smooth over the graphs.

Localization is crucial for various applications, this includes resource coordination in small and ultra-small cells, as well as the whole range of Location Based Service (LBS). Multilateration is a localization technique that is based on distance measurements between multiple reference nodes and a target node. This paper introduces a multilateration localization approach that uses Singular Value Decomposition (SVD) for 3D indoor positioning. It also provides a mathematical multilateration formulation which considers the coordinates of the reference nodes and the relative distance between transmitting nodes. In practical deployments, the relative distance can be estimated using RSSI; we apply Kalman filtering to the RSSI measurements aiming to get a more accurate RSSI value. The approach is complemented by using two selection methods which help chosing the best nodes for multilateration computation. The paper concludes with a discussion of the experimental evaluation results obtained.

In this paper, we present the design and implementation of a prototype system of Smart Parking Services based on Wireless Sensor Networks (WSNs) that allows vehicle drivers to effectively find the free parking places. The proposed scheme consists of wireless sensor networks, embedded web-server, central web-server and mobile phone application. In the system, low-cost wireless sensors networks modules are deployed into each parking slot equipped with one sensor node. The state of the parking slot is detected by sensor node and is reported periodically to embedded web-server via the deployed wireless sensor networks. This information is sent to central web-server using Wi-Fi networks in real-time, and also the vehicle driver can find vacant parking lots using standard mobile devices.

BACKGROUND: The construction of interaction networks between proteins is central to understanding the underlying biological processes. However, since many useful relations are excluded in databases and remain hidden in raw text, a study on automatic interaction extraction from text is important in bioinformatics field. RESULTS: Here, we suggest two kinds of kernel methods for genic interaction extraction, considering the structural aspects of sentences. First, we improve our prior dependency kernel by modifying the kernel function so that it can involve various substructures in terms of (1) e-walks, (2) partial match, (3) non-contiguous paths, and (4) different significance of substructures. Second, we propose the walk-weighted subsequence kernel to parameterize non-contiguous syntactic structures as well as semantic roles and lexical features, which makes learning structural aspects from a small amount of training data effective. Furthermore, we distinguish the significances of parameters such as syntactic locality, semantic roles, and lexical features by varying their weights. CONCLUSIONS: We addressed the genic interaction problem with various dependency kernels and suggested various structural kernel scenarios based on the directed shortest dependency path connecting two entities. Consequently, we obtained promising results over genic interaction data sets with the walk-weighted subsequence kernel. The results are compared using automatically parsed third party protein-protein interaction (PPI) data as well as perfectly syntactic labeled PPI data.

We present a new distributed genetic algorithm that can be used to extract useful information from distributed, large data over the network. The main idea of the proposed algorithm is to determine how many and which individuals move between subpopulations at each site adaptively. In addition, we present a method to help individuals from other subpopulations not be weeded out but adapt to the new subpopulation. We apply our distributed genetic algorithm to the feature subset selection task which has been one of the active research topics in machine learning. We used six data sets from UCI Machine Learning Repository to compare the performance of our approach with that of the single, centralized genetic algorithm. As a result, the proposed algorithm produced better performance than the single genetic algorithm in terms of the classification accuracy with the feature subsets.

We introduce a new adaptive genetic method for AVT generation, MCM-AVT-Learner. The MCM-AVT-Learner imports the mutation and crossover matrices which makes effective use of the fitness ranking and loci statistics information. The suggested method is not only parameter-free, but also capable of producing high quality AVTs. We describe experiments on several complete and missing benchmark data sets that compare the performance of AVT-DTL using the reslut AVTs of the MCM-AVT-Learner and existing AVT learning algorithms. Results show that the AVTs generated by MCM-AVT-Learner are competitive with human-generated AVTs or AVTs generated by HAC-AVT-Learner and GA-AVT-Learner in terms of classification accuracy and the compactness of the classifier.

We present a new ensemble learning method that employs a set of regional classifiers, each of which learns to handle a subset of the training data. We split the training data and generate classifiers for different regions in the feature space. When classifying new data, we apply a weighted voting among the classifiers that include the data in their regions. We used 10 datasets to compare the performance of our new ensemble method with that of single classifiers as well as other ensemble methods such as bagging and Adaboost. As a result, we found that the performance of our method is comparable to that of Adaboost and bagging when the base learner is C4.5. In the remaining cases, our method outperformed the benchmark methods.

‘Bioinformatics—From Genomes to Therapies’, edited by Thomas Lengauer, is big collection of contributions that includes the work of 91 authors arranged in 45 chapters, 11 parts and three volumes of 1814 pages. As an expanded sequel of the author's previous publication, ‘Bioinformatics—From Genomes to Drugs’, the book includes diverse topics in bioinformatics focusing on the challenges in understanding, diagnosing and curing of diseases. Remembering aforementioned focus on the design, development and therapies of drugs for diseases, the three volumes are organized to help understand the building blocks of sequences and structures (Volume 1), the inner workings of molecular interactions (Volume 2) and the integrated picture of molecular function (Volume 3), respectively. Each volume in turn consists of several parts, each of which consists of several chapters under the same umbrella of a specific topic. References are provided at the end of each chapter giving both beginners and more experienced researchers additional information on the chapter. In addition, there is a global index that includes an extensive list of definitions, concepts and methods in bioinformatics addressed throughout the book. To sum up, Volume 1 is composed of 15 chapters in four parts: Introduction (Part 1), Sequencing Genomes (Part 2), Sequence Analysis (Part 3) and Molecular Structure Prediction (Part 4). Volume 2 is composed of 13 chapters in three parts: Analysis of Molecular Interactions (Part 5), Molecular Networks (Part 6) and Analysis of Expression Data (Part 7). Volume 3 is composed of 17 chapters in four parts: Protein Function Prediction (Part 8), Comparative Genomics and Evolution of Genomes (Part 9), Basic Bioinformatics Technologies (Part 10) and Outlook (Part 11). The book begins with Part 1 consisting of an introductory chapter (Chapter 1) that explains the molecular basics of a disease and molecular approach to its cure (including protein targeting, genomics and proteomics, information on genes and proteins, drug development, therapy optimization). Chapter 1 also contains a nice summary on the organization of the book, from which this review is written. Part 2 consists of a single chapter (Chapter 2) that discusses the assembly strategies for genome sequencing with algorithmic issues followed by examples of some of the existing assemblers. Part 3 comprises Chapters 3–8 on sequence analysis. Chapter 3 includes the pairwise and multiple sequence alignment techniques and the sequence search in databases. Chapter 4 explains the phylogeny reconstruction approaches for ancestral inference by the tree reconstruction from the leaves, finding the optimal tree and then generating split networks from the trees. Chapters 5–7 start with introductory material followed by common methodologies for finding the protein-coding genes, for analyzing the regulatory regions and for finding the repeats in genomes, respectively. Chapter 8 ends Part 3 touching on the analysis of genome rearrangements with various methods (e.g. distance-based, maximum parsimony, maximum likelihood). Part 4 deals with the molecular structure prediction in Chapter 9 through 15. Chapter 9 explains the hard task of three-dimensional protein structure prediction and describes simplified features of the task including the secondary structures, transmembrane regions, solvent accessibility, inter-residue contacts, flexible and intrinsically disordered regions and protein domains. Chapter 10 covers the homology modeling in biology and medicine (i.e. structure prediction based on the sequence alignment with template proteins) provided with modeling methods and results, followed by Chapter 11 that deals with the protein fold recognition task based on distant homologs, using a variety of computational models (e.g. Hidden Markov Models, Support Vector Machines). Chapter 12 explores the De Novo structure prediction method (as a method that does not rely on homology between the query and the prototype sequences), and its application in systems biology. Chapter 13 is an introductory chapter on structural genomics that attempts to predict protein structures over whole proteomes and combines the experimental as well as homology-modeling based approaches. Chapters 14 and 15 conclude Part 4 by discussing the secondary and tertiary structure prediction of RNA, respectively. In Part 5, molecular interactions are discussed in Chapter 16 through 19, focusing on the interactions between a drug and its target protein for drug therapy. Chapter 16 explores the docking and scoring methods for drug design, explaining how to dock ligands into protein-binding sites and how to assemble new ligands inside the binding site. Chapter 17 covers a variety of approaches to protein–protein and protein–DNA docking (e.g. Correlation, Monte Carlo techniques). Chapters 18 and 19 discuss lead identification and optimization (for drug discovery) using virtual screening and combination of relevant features (or biological variables), with the aid of data mining and automated decision-making models. Part 6 covers molecular networks in four chapters, in order to understand the interactions between diseases and their therapies. Chapter 20 exhibits various approaches for modeling and simulating metabolic networks statically as well as dynamically, considering the red bold cell metabolism as an example. Chapter 21 introduces the motivation and requirements of gene regulatory networks and discusses methods for the inference of the networks. Chapter 22 explains cell signaling networks and the methods for their analysis (e.g. Boolean, Bayesian networks). Chapter 23 concludes Part 6 by bringing about a medical application, the dynamics of interaction between the viruses and the host cells. Part 7 comprises Chapter 24 through 28, focusing on the analysis of expression data. Chapter 24 contains an overview on the technologies and applications in DNA microarrays. Chapter 25 explains the low-level details of processes in microarray experiments. Chapters 26 and 27 touch on the classification of patients and genes, respectively, and exhibit a wide spectrum of classification and clustering algorithms that have been developed and used commonly in machine learning and data mining research communities. Chapter 28 introduces a relatively new area, Proteomics, beyond the genome analysis to deal with the diverse and dynamic characteristics of proteins. Part 8 discusses the prediction of protein functions in eight chapters. Chapter 29 introduces diverse ontologies in molecular biology that need to be clarified prior to the prediction. Chapter 30 explains techniques for inferring protein functions from sequences and describes representative databases, as well. Chapter 31 explores various approaches to the analysis of protein interaction networks, based on different types of data (e.g. sequences, structures, textual information). Chapters 32 and 33 again focus on the inference of protein functions, based on genomic context (i.e. genomes, genes, gene arrangements) and protein structures, respectively. Chapter 34 deals with text mining on protein functions such as information retrieval and extraction, question answering and natural language generation. Chapter 35 discusses integration of different data such as features, classifiers and networks for protein function prediction. Chapter 36 finally discusses methods for predicting druggability of proteins in drug design considering the suitability of their shapes in binding. Part 9 comprises Chapter 37 through 41 and discusses the comparative genomics and the evolution of genomes, focusing on the analysis of relationships and differences between genomes. Chapter 37 provides the definition, motivation and technologies of comparative genomics in a bid to figure out what we can learn from different genomes of individuals of the same species. Chapter 38 deals with the association studies of diseases to understand the genetic differences among people on the susceptibility of the disease, and provides related statistical methods. Chapter 39 starts with introductory material on pharmacogenetics and pharmacogenomics for personalized drugs, and provides associated data and tools. Chapter 40 focuses on the evolution of drug resistance in HIV, providing introductory material as well as techniques, resources and issues in the domain. Chapter 41 is another chapter focusing on a specific task, analysis of the evolution of infectious bacteria. Part 10 addresses some of the basic technologies in bioinformatics in three chapters. Chapter 42 touches on the issue of integrating diverse biological databases and explains the data models, integration methods and implementation technologies. Chapter 43 handles the data visualization, showing many methods for different types of data. Chapter 44 concludes Part 10 by delivering the issues and technologies of distributed computing in bioinformatics. The last part (Part 11) is an outlook in bioinformatics. It includes a concluding chapter (Chapter 45) that summarizes the field and enumerates future research areas in bioinformatics and computational biology. With the diverse but integrated contents provided for the molecular analysis of diseases and their therapies, the book can be a valuable asset to the researchers from the computer science, mathematics, statistics and biological sciences, in performing interdisciplinary research in bioinformatics in general and in pharmaceutics and molecular medicine in particular. Especially for computer scientists, the book provides relevant knowledge and addresses research issues in the biological domain, which makes it possible for them to accomplish significant achievements based on their computational and algorithmic backgrounds in computer science.

MOTIVATION: Automatic knowledge discovery and efficient information access such as named entity recognition and relation extraction between entities have recently become critical issues in the biomedical literature. However, the inherent difficulty of the relation extraction task, mainly caused by the diversity of natural language, is further compounded in the biomedical domain because biomedical sentences are commonly long and complex. In addition, relation extraction often involves modeling long range dependencies, discontiguous word patterns and semantic relations for which the pattern-based methodology is not directly applicable. RESULTS: In this article, we shift the focus of biomedical relation extraction from the problem of pattern extraction to the problem of kernel construction. We suggest four kernels: predicate, walk, dependency and hybrid kernels to adequately encapsulate information required for a relation prediction based on the sentential structures involved in two entities. For this purpose, we view the dependency structure of a sentence as a graph, which allows the system to deal with an essential one from the complex syntactic structure by finding the shortest path between entities. The kernels we suggest are augmented gradually from the flat features descriptions to the structural descriptions of the shortest paths. As a result, we obtain a very promising result, a 77.5 F-score with the walk kernel on the Language Learning in Logic (LLL) 05 genic interaction shared task. AVAILABILITY: The used algorithms are free for use for academic research and are available from our Web site http://mllab.sogang.ac.kr/ approximately shkim/LLL05.tar.gz.

With the proliferation of the Internet and the huge amount of data it transfers, text summarization is becoming more important. We present an approach to the design of an automatic text summarizer that generates a summary by extracting sentence segments. First, sentences are broken into segments by special cue markers. Each segment is represented by a set of predefined features (e.g. location of the segment, average term frequencies of the words occurring in the segment, number of title words in the segment, and the like). Then a supervised learning algorithm is used to train the summarizer to extract important sentence segments, based on the feature vector. Results of experiments on U.S. patents indicate that the performance of the proposed approach compares very favorably with other approaches (including Microsoft Word summarizer) in terms of precision, recall, and classification accuracy.

Constructive learning algorithms offer an attractive approach for the incremental construction of near-minimal neural-network architectures for pattern classification. They help overcome the need for ad hoc and often inappropriate choices of network topology in algorithms that search for suitable weights in a priori fixed network architectures. Several such algorithms are proposed in the literature and shown to converge to zero classification errors (under certain assumptions) on tasks that involve learning a binary to binary mapping (i.e., classification problems involving binary-valued input attributes and two output categories). We present two constructive learning algorithms MPyramid-real and MTiling-real that extend the pyramid and tiling algorithms, respectively, for learning real to M-ary mappings (i.e., classification problems involving real-valued input attributes and multiple output classes). We prove the convergence of these algorithms and empirically demonstrate their applicability to practical pattern classification problems. Additionally, we show how the incorporation of a local pruning step can eliminate several redundant neurons from MTiling-real networks.

Bioinformatics and related applications call for efficient algorithms for knowledge-intensive learning and data-driven knowledge refinement. Knowledge based artificial neural networks offer an attractive approach to extending or modifying incomplete knowledge bases or domain theories. We present results of experiments with several such algorithms for data-driven knowledge discovery and theory refinement in some simple bioinformatics applications. Results of experiments on the ribosome binding site and promoter site identification problems indicate that the performance of KBDistAl and Tiling-Pyramid algorithms compares quite favorably with those of substantially more computationally demanding techniques.

Multi-layer networks of threshold logic units (TLU) offer an attractive framework for the design of pattern classification systems. A new constructive neural network learning algorithm (DistAl) based on inter-pattern distance is introduced. DistAl constructs a single hidden layer of hyperspherical threshold neurons. Each neuron is designed to determine a cluster of training patterns belonging to the same class. The weights and thresholds of the hidden neurons are determined directly by comparing the inter-pattern distances of the training patterns. This offers a significant advantage over other constructive learning algorithms that use an iterative (and often time consuming) weight modification strategy to train individual neurons. The individual clusters (represented by the hidden neurons) are combined by a single output layer of threshold neurons. The speed of DistAl makes it a good candidate for datamining and knowledge acquisition from large datasets. The paper presents results of experiments using several artificial and real-world datasets. The results demonstrate that DistAl compares favorably with other learning algorithms for pattern classification.