Self-learning Approach Research Articles

Minimally-Biased Metadynamics Method to Sample Conformational Ensembles Compatible with Experimental Measurements

A primary goal in computational biophysics is to harness experimental measurements to obtain information on the structure and dynamics of biomolecules. However, most biophysical techniques such as NMR and EPR spectroscopy provide signals that arise from an ensemble of multiple molecular conformations. Thus, it is typically not straightforward to extract detailed structural information from the experimental data. A possible strategy is to bias the conformational sampling obtained in a molecular dynamics simulations in reference to the experimental data, under the so-called maximum entropy principle. Recent practical formulations of this approach involve simulations carried out over multiple replicas or iterative ensemble-correction procedures based on the determination of several (Lagrange) parameters. Here, we present an alternative, self-learning approach to sample molecular ensembles compatible with experimental data with the minimal possible bias on the simulation trajectories. The method does not require multiple replicas and is based on adding an adaptive bias potential during the simulation that discourages the sampling of conformations that are not consistent with the experimental measurements. To illustrate this approach, we applied this novel simulation technique to spin-labeled T4-lysozyme, targeting a set of spin-spin distance distributions measured by DEER/EPR spectroscopy. We show how the proposed method is able to efficiently sample the experimental distance distributions without altering uncorrelated degrees of freedom. We anticipate that this new simulation approach will be widely useful to obtain conformational ensembles compatible with diverse types of experimental measurements of biomolecular dynamics.

Anomaly Extraction and Mitigation using Efficient-Web Miner Algorithm

Today network security, uptime and performance of network are important and serious issues in computer network. Anomaly is deviation from normal behavior affecting network security. Anomaly Extraction is identification of unusual flow from network, which is need of network operator. Anomaly extraction aims to automatically find the inconsistencies in large set of data observed during an anomalous time interval. Extracted anomalies will be important for root cause analysis, network forensics, attack mitigation and anomaly modeling. Frequent pattern mining technique namely Efficient-Web Miner Algorithm will be used to generate the set of association rules applied on metadata. Using network traffic log data, algorithms effectively finds the flow associated with the anomalous event(s). EfficientWeb Miner Algorithm triggers a very small number of false positives. EfficientWeb Miner has much better performance in terms of time and space complexity than Apriori Algorithm and its variations like Apriori All algorithm.for large data sets This anomaly extraction method significantly reduces the time needed for analyzing alarms, making anomaly detection systems more practical, simple and realistic. System makes an effort to mitigate the anomaly so detected without human intervention. Proposed system provides human overrides in mitigation process and inculcates self-learning approach which is advantageous.

New artificial neural networks for true triaxial stress state analysis and demonstration of intermediate principal stress effects on intact rock strength

Simulations are conducted using five new artificial neural networks developed herein to demonstrate and investigate the behavior of rock material under polyaxial loading. The effects of the intermediate principal stress on the intact rock strength are investigated and compared with laboratory results from the literature. To normalize differences in laboratory testing conditions, the stress state is used as the objective parameter in the artificial neural network model predictions. The variations of major principal stress of rock material with intermediate principal stress, minor principal stress and stress state are investigated. The artificial neural network simulations show that for the rock types examined, none were independent of intermediate principal stress effects. In addition, the results of the artificial neural network models, in general agreement with observations made by others, show (a) a general trend of strength increasing and reaching a peak at some intermediate stress state factor, followed by a decline in strength for most rock types; (b) a post-peak strength behavior dependent on the minor principal stress, with respect to rock type; (c) sensitivity to the stress state, and to the interaction between the stress state and uniaxial compressive strength of the test data by the artificial neural networks models (two-way analysis of variance; 95% confidence interval). Artificial neural network modeling, a self-learning approach to polyaxial stress simulation, can thus complement the commonly observed difficult task of conducting true triaxial laboratory tests, and/or other methods that attempt to improve two-dimensional (2D) failure criteria by incorporating intermediate principal stress effects.

Adaptive Multi-Sensor Perception for Driving Automation in Outdoor Contexts

In this research, adaptive perception for driving automation is discussed so as to enable a vehicle to automatically detect driveable areas and obstacles in the scene. It is especially designed for outdoor contexts where conventional perception systems that rely on a priori knowledge of the terrain's geometric properties, appearance properties, or both, is prone to fail, due to the variability in the terrain properties and environmental conditions. In contrast, the proposed framework uses a self-learning approach to build a model of the ground class that is continuously adjusted online to reflect the latest ground appearance. The system also features high flexibility, as it can work using a single sensor modality or a multi-sensor combination. In the context of this research, different embodiments have been demonstrated using range data coming from either a radar or a stereo camera, and adopting self-supervised strategies where monocular vision is automatically trained by radar or stereo vision. A comprehensive set of experimental results, obtained with different ground vehicles operating in the field, are presented to validate and assess the performance of the system.

Semi-supervised learning of causal relations in biomedical scientific discourse.

BackgroundThe increasing number of daily published articles in the biomedical domain has become too large for humans to handle on their own. As a result, bio-text mining technologies have been developed to improve their workload by automatically analysing the text and extracting important knowledge. Specific bio-entities, bio-events between these and facts can now be recognised with sufficient accuracy and are widely used by biomedical researchers. However, understanding how the extracted facts are connected in text is an extremely difficult task, which cannot be easily tackled by machinery.ResultsIn this article, we describe our method to recognise causal triggers and their arguments in biomedical scientific discourse. We introduce new features and show that a self-learning approach improves the performance obtained by supervised machine learners to 83.47% for causal triggers. Furthermore, the spans of causal arguments can be recognised to a slightly higher level that by using supervised or rule-based methods that have been employed before.ConclusionExploiting the large amount of unlabelled data that is already available can help improve the performance of recognising causal discourse relations in the biomedical domain. This improvement will further benefit the development of multiple tasks, such as hypothesis generation for experimental laboratories, contradiction detection, and the creation of causal networks.

An EPR-based self-learning approach to material modelling

In this paper an EPR-based self-learning method is presented for modelling the constitutive behaviour of materials using evolutionary polynomial regression (EPR). The proposed approach takes advantage of the rich stress–strain data buried in non-homogenous structural tests. The load–deformation data collected from experiment are used to iteratively train EPR-based material model using finite element simulations of the structural test. Two numerical examples are presented to illustrate the application of the proposed approach. It is shown that the EPR model gradually improves during the self-learning training and provides accurate prediction for the constitutive behaviour of the material.

A Self-Learning Approach to Single Image Super-Resolution

Learning-based approaches for image super-resolution (SR) have attracted the attention from researchers in the past few years. In this paper, we present a novel self-learning approach for SR. In our proposed framework, we advance support vector regression (SVR) with image sparse representation, which offers excellent generalization in modeling the relationship between images and their associated SR versions. Unlike most prior SR methods, our proposed framework does not require the collection of training low and high-resolution image data in advance, and we do not assume the reoccurrence (or self-similarity) of image patches within an image or across image scales. With theoretical supports of Bayes decision theory, we verify that our SR framework learns and selects the optimal SVR model when producing an SR image, which results in the minimum SR reconstruction error. We evaluate our method on a variety of images, and obtain very promising SR results. In most cases, our method quantitatively and qualitatively outperforms bicubic interpolation and state-of-the-art learning-based SR approaches.

A self-learning approach to improving service quality in outsourcing of engineering design using operational data

Managing service quality in outsourcing requires a holistic approach to managing knowledge. This need is more pronounced in case of outsourcing of high-end tasks such as engineering designs. As complex tasks are carried out large amounts of multi-structured transactional data are captured during service delivery, more often appearing as text. This qualitative operational data has rich knowledge embedded in it. This paper aims to demonstrate a way of extracting knowledge from such operational data for improving service quality. The study uses simulation as a method of inductive research. Simulation model of a self-learning system for extracting knowledge from operational data are created. The proposed artificial intelligence system integrates natural language processing and rule-based reasoning for knowledge creation. Finally, with a view to demonstrate the potential of the proposed system, a real prototype industry application is described.

Analysis of asymmetric driving behavior using a self-learning approach

This paper presents a self-learning Support Vector Regression (SVR) approach to investigate the asymmetric characteristic in car-following and its impacts on traffic flow evolution. At the microscopic level, we find that the intensity difference between acceleration and deceleration will lead to a ‘neutral line’, which separates the speed-space diagram into acceleration and deceleration dominant areas. This property is then used to discuss the characteristics and magnitudes of microscopic hysteresis in stop-and-go traffic. At the macroscopic level, according to the distribution of neutral lines for heterogeneous drivers, different congestion propagation patterns are reproduced and found to be consistent with Newell’s car following theory. The connection between the asymmetric driving behavior and macroscopic hysteresis in the flow-density diagram is also analyzed and their magnitudes are shown to be positively related.

Database Development to Enhance Procedures for Open Ended Laboratory

In the globalization era and fast development in areas of science and engineering, the development of computer-based technology has become one of the contributors to national development. Biochemical and Chemical Engineering Laboratory, Department of Chemical and Process Engineering, Faculty of Engineering and Built Environment, UKM has applied the self-learning approach through Open Ended laboratory. To continue with this learning method, such experiments should be done repeatedly for a decision, other than a lot of time had been provided to choose an appropriate title. A study on the development of database for open ended laboratory procedure was carried out to overcome this problem. In this study, the software used is Microsoft Visual Basic (VB) version 6.0 with Malay Language as the main language, VB application which is very suitable for developing such a database. This database was produced by providing a home, open-ended laboratory site and page title list to choose the titles of the open-ended laboratory. Based on survey results, the database for open-ended laboratory procedures was developed which provide information on the topics of open-ended laboratory was conducted as well as cordial user. Finally, comparisons are made between the resulting databases with existing database.

Self-Learning is the Future: A New Paradigm for the 21st Century

Self-Learning method is an individualized method of learning. The face-to-face teaching is disappearing and distance mode of education is becoming popular. Self-learning modules are designed where the learner is free to chose what to learn, how to learn, when to learn and where to learn. This flexibility is an importance characteristic in open learning process. With the advent of information technology for communication, the other modes of instructions are slowly disappearing. The learner is getting accustomed more and more to non-formal mode of education thereby shifting the preference to self-learning methods. Even most of the print materials that come into the market are meant for self-learning. The learner is at an advantage to use this form of non-formal mode of education using self-paced learning materials. The teaching-learning-evaluation all happens without face-to-face interaction. There is no fixed time and no fixed number of classes. To make this learning process effective, it is essential for a Trainer to have the conceptual knowledge of self-learning approach and also the design of self-learning modules and materials. The Trainer should also look at the three domains of learning viz., cognitive, affective and psychomotor and the emphasis required while designing the learning resource material. This paper discusses the features of self-learning approach and design of self-learning modules and material which could be used effectively in non-formal mode of education. The paper also highlights the use of Computer Assisted Learning (CAL) packages used for self-learning purpose.

Self-learning classification of radar features for scene understanding

Autonomous driving is a challenging problem in mobile robotics, particularly when the domain is unstructured, as in an outdoor setting. In addition, field scenarios are often characterized by low visibility as well, due to changes in lighting conditions, weather phenomena including fog, rain, snow and hail, or the presence of dust clouds and smoke. Thus, advanced perception systems are primarily required for an off-road robot to sense and understand its environment recognizing artificial and natural structures, topology, vegetation and paths, while ensuring, at the same time, robustness under compromised visibility. In this paper the use of millimeter-wave radar is proposed as a possible solution for all-weather off-road perception. A self-learning approach is developed to train a classifier for radar image interpretation and autonomous navigation. The proposed classifier features two main stages: an adaptive training stage and a classification stage. During the training stage, the system automatically learns to associate the appearance of radar data with class labels. Then, it makes predictions based on past observations. The training set is continuously updated online using the latest radar readings, thus making it feasible to use the system for long range and long duration navigation, over changing environments. Experimental results, obtained with an unmanned ground vehicle operating in a rural environment, are presented to validate this approach. A quantitative comparison with laser data is also included showing good range accuracy and mapping ability as well. Finally, conclusions are drawn on the utility of millimeter-wave radar as a robotic sensor for persistent and accurate perception in natural scenarios.

An improved approach of self-organising fuzzy neural network based on similarity measures

This paper proposes a self-learning approach to automatically generate an accurate and compact fuzzy neural network. The proposed algorithm is an integrated approach including addition, pruning and merging strategies. The merging strategy is based on a similarity measure of membership functions. Such an integrated approach has the advantages of reducing model overfitting, enhancing interpretability and improving computing effectiveness. Benchmark examples are presented for comparative analysis of model accuracy and compactness.

Novel Nature Inspired Techniques in Medical Data Mining

In this work we have studied, evaluated and proposed different swarm intelligence techniques for mining information from loosely structured medical textual records with no a priori knowledge (a large dataset). The output of this task is a set of ordered/nominal attributes suitable for rule discovery mining.Information mining from textual data becomes a very challenging task when the structure of the text record is loose without any rules. The task becomes even harder when natural language is used and no a priori knowledge is available.First, classical approaches such as basic statistic approaches, single and multiple word frequency analysis, etc., have been used to simplify the textual data and provide an overview of the data. Finally, an ant-inspired self-learning approach has been used to automatically provide a simplified dominant structure, presenting structure of the records in the human readable form that can be further utilized in the mining process as it describes the vast majority of the records.Note that this project is an ongoing process (and research) and new data are irregularly received from the medical facility, justifying the need for robust and fool-proof algorithms.

Optimal dynamic pricing strategies for high-occupancy/toll lanes

This paper proposes a self-learning approach to determine optimal pricing strategies for high-occupancy/toll lane operations. The approach learns recursively motorists’ willingness to pay by mining the loop detector data, and then specifies toll rates to maximize the freeway’s throughput while ensuring a superior travel service to the users of the toll lanes. In determination of the tolls, a multi-lane hybrid traffic flow model is used to explicitly consider the impacts of the lane-changing behaviors before the entry points of the toll lanes on throughput and travel time. Simulation experiments are conducted to demonstrate and validate the proposed approach, and provide insights on when to convert high-occupancy lanes to toll lanes.

Photon-number squeezing in a solitonlike Raman Stokes component during propagation of ultrashort pulses in a microstructure fiber

We discuss the characteristics of photon-number squeezing generated in a spectrally filtered Raman Stokes pulse obtained from a nonlinear fiber. The Raman Stokes pulse generated from a microstructure fiber with a femtosecond input laser pulse exhibits far higher photon-number squeezing. We studied the quantum correlation established among the frequency modes in the Raman Stokes pulse by using numerical calculations and experiments. To design an intra-Stokes pulse quantum correlation through the optical nonlinearities of a fiber to obtain higher photon-number squeezing, we shaped the input laser pulse by using a self-learning closed-loop control approach.

DEVELOPING A LARGE-SCALE URBAN DECISION SUPPORT SYSTEM

A large-scale Decision Support System (DSS) is being developed and will be applied for Beijing city in China. The main purpose is to be able to propose best suitable measures for a given (either recurrent or non-recurrent) traffic situation, and to apply it to a real-life traffic management, with focus on the application around the Olympics Area. A major issue for operational management is to be able fast to recognize primary problems and to be quick to recommend/retrieve corresponding solutions. This paper proposes a novel self-learning approach using conjointly expert knowledge-based choice and case-based reasoning. Key aspects to support such process include: (a) problem identification that is based on a mesoscopic large-scale network dynamic simulation; (b) measures that have been successfully implemented in a priori cases would serve as new initial scenarios to the new situations, and (c) measure evaluation that can be performed according to performance indictors. Effective scenarios (measure to problem) are stored into KBEST (knowledge-based expert system) and made available for offline and online calls. System building and a calibration process are being followed, and an implementation of such system to an incident management and route guidance is foreseen and being designed.

Book Review: Logistic regression A self-learning approach, 2nd edition

Book Review: Logistic regression A self-learning approach, 2nd edition

Gross anatomy dissections and self-directed learning in medicine

The Medical School of the University of Castilla-La Mancha (UCLM, Albacete Spain) was launched in 1998 and is the most recent one in Spain. Teaching is based on small groups of students (20-25 students/group). An objective-oriented self-learning approach provides maximal autonomy and independence in the achievement of objectives by the students in close association with academic staff. Gross Anatomy courses take place in the first and second years. The one in the first year is a single 10-credit course, where one credit equals 10 hr of teaching activity. In the second year, Anatomy and Embryology are integrated with Physiology and Histology, and comprise 70 credits altogether. In addition, all students carry out two mandatory gross anatomical dissections per year, in groups of three students, to allow direct handling of human anatomical material. Students are provided with handouts containing general instructions on how to perform the dissection and the structures (items) that they must expose in a given period of time (4 hr). Afterward, a Faculty member checks the number of items demonstrated and the quality of the dissection. Each group submits a written report that contributes to the final score. We evaluated the number of items shown in each of two consecutive dissections for first and second year medical students. The data obtained indicate that students engaged in self-directed learning through small groups working with Faculty staff are able to self-improve their anatomical skills.

In-phase implies large likelihood for independent codon model: distinguishing coding from non-coding sequences

It is proven that under the independent codon model, the likelihood of a DNA coding sequence read according to the correct frame is asymptotically larger than that read with an incorrect frame. Based on this proposition, a single set of probabilities of the codon usage is enough for discriminating the six frames of coding sequences under the independent codon model. The direct coding sequence of Escherichia coli genome is taken as an example to examine the codon independency by using the mutual information and χ 2 analysis. The contrast between the coding frame and the two offset frames is evident. A self-learning approach for generating training set is proposed to estimate probability parameters.