Machine Learning Classification Tools Research Articles

Development of a Fault Detection and Localization Model for a Water Distribution Network

Water distribution networks are complex systems that aid in the delivery of water to residential and non-residential areas. However, the networks can be affected by different types of faults, which could lead to the wastage of treated water. As such, there is a need to develop a reliable leakage detection and localization system that can detect leak occurrences in the network. This study, using a simulated dataset from EPANET, presents the application of supervised machine learning classifiers for leak detection and localization in the water distribution network of the University of Port Harcourt Choba campus. The study compared three machine learning classification tools that are used in pattern recognition analysis: the support vector machine, k-nearest neighbor, and artificial neural network. The robustness and effectiveness of the proposed approach are compared with those of the performance of the classifiers for leakage detection in the network of the case study. The results show that the support vector machine performs the best, with 79% accuracy, while the respective accuracies for the remaining classifiers are 70% for the k-nearest neighbor and 61% for the artificial neural networks. The high accuracy demonstrated by the models shows that they are able to detect and address issues relating to fault detection in a water distribution network. This model could provide a leakage detection system to be applied to buildings for the efficient management of water in their networks.

Assessment of subvisible particles in biopharmaceuticals with image feature extraction and machine learning

An image classification tool was developed to classify subvisible particles, namely silicone oil (SO) and non-silicone oil (NSO; protein aggregate, rubber closure, and air bubble) particles, present in biopharmaceuticals using feature extraction in FlowCam® images, and the outcomes were validated with six machine learning (ML) classifiers. The image classification tool set at specific configuration: condition 1 − CDEP {(compactness, diameter, elongation, perimeter) = (8.06, 10.44, 13.30, 27.20)} identified SO, while the configuration set at condition 2 – CDEL {(compactness, diameter, elongation, length) = (3.14, 24.48, 1.97, 4.28)} detected NSO. The classification tool was particularly useful in detecting the release of SO after exposure to the stress sources. Additionally, the morphological features-based classification tool (p < 0.05) enhanced the predictive accuracy of the ML classification tools (≥97.2 %). Specifically, CNN (100 %) outperformed naïve Bayes (99.3 %), linear discriminant analysis (98.4 %), artificial neural network (98.1 %), support vector machine (SVM 97.2 %). Bootstrap forest was excluded because it failed to classify SO in a large dataset. The developed classification tool could be an alternative in classifying the image datasets without the burden of complex ML tools. Such image-based classification tool can be computationally economical solution in quality control of the protein formulations.

Characterization of the shape of aggregates using image analysis and machine learning classification tools

ABSTRACT Engineering applications including pervious concrete require effective packing of aggregates to optimize strength. Size and shape distribution of aggregates significantly affect the performance. Computational methods numerically represent shape of aggregates, from image analysis, the effectiveness of has not been compared and verified. This study aims to analyse the representability of shape aspects of aggregates by different computational methods. Crushed aggregates were grouped into 5 clusters, and each group was milled in a Los Angeles machine for different degrees (0–2000) to induce morphological changes on the aggregates. Aggregates ranging from 5 to 30mm in diameter were obtained (7191 in total). imageJTM, was used to compute dimensions and shape factors of aggregates from 14 computational methods. Statistical tests, Pearson’s Correlations and Principal Components Analysis and machine learning classification tools, Decision-tree, Random-Forest, Naïve-Bayes, Support-Vector-Machines, K-Nearest-Neighbours and Perceptron were employed to assess. In conclusion, no shape factor could be singularly used to numerically represent the morphological changes on aggregate particles but a combination of shape factors is required. Data matrix had three primary dimensions. Combination of Circularity, Kumbrein-Solidity and Barksdale-Shape-Factor yield best representation of aggregate shape. Regression Tree classification method had the highest accuracy (0.9) in classifying milled and unmilled aggregates.

ARTIFICIAL INTELLIGENCE TO DEAL WITH THE POST COVID-19 FRACTAL DYNAMICS LINKED WITH ECONOMY

Data-based studies have always provided useful insights of the research problem, provided that correct statistical modeling and inference strategies are adopted. For bi-directional studies or longitudinal datasets, it is always difficult to analyze the dependence of variables and their impact on the key features. With the advancement in the fields of data science and the applied mathematical modeling, these difficulties are well resolved. The machine learning algorithms can help to streamline the data-based studies in a robust manner. During this research, the COVID-19 data are analyzed with the aid of machine learning classification tools to identify the predictors, directly influenced by the pandemic. The impact of COVID-19 on the world’s economy can be better interpreted with the aid of data-based research. The data linked to the unemployment rates, during the frequent waves of COVID-19, are extracted from different sources and analyzed during this research for better forecasting measures. The nonlinear dynamical analysis can be visualized with the aid of the 2D fractal pattern generation approach. Thus the current research is an attempt to connect the outcomes of the classification analysis to the 2D fractal generation for better visual interpretation of lengthy large datasets.

Characteristics and outcomes of SARS-COV 2 critically ill patients after emergence of the variant of concern 20H/501Y.V2: A comparative cohort study.

There are currently no data regarding characteristics of critically ill patients with Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) variant of concern (VOC) 20H/501Y.V2. We therefore aimed to describe changes of characteristics in critically ill patients with Covid-19 between the first and the second wave when viral genome sequencing indicated that VOC was largely dominant in Mayotte Island (Indian Ocean). Consecutive patients with Covid-19 and over 18 years admitted in the unique intensive care unit (ICU) of Mayotte during wave 2 were compared with an historical cohort of patients admitted during wave 1. We performed a LR comparing wave 1 and wave 2 as outcomes. To complete analysis, we built a Random Forest model (RF), that is, a machine learning classification tool- using the same variable set as that of the LR. We included 156 patients, 41 (26.3%) and 115 (73.7%) belonging to the first and second waves respectively. Univariate analysis did not find difference in demographic data or in mortality. Our multivariate LR found that patients in wave 2 had less fever (absence of fever aOR 5.23, 95% confidence interval (CI) 1.89–14.48, p = .001) and a lower simplified acute physiology score (SAPS II) (aOR 0.95, 95% CI 0.91–0.99, p = .007) at admission; at 24 hours, the need of invasive mechanical ventilation was higher (aOR 3.49, 95% CI 0.98–12.51, p = .055) and pO2/FiO2 ratio was lower (aOR 0.99, 95 % CI 0.98–0.99, p = .03). Patients in wave 2 had also an increased risk of ventilator-associated pneumonia (VAP) (aOR 4.64, 95% CI 1.54–13.93, p = .006). Occurrence of VAP was also a key variable to classify patients between wave 1 and wave 2 in the variable importance plot of the RF model. Our data suggested that VOC 20H/501Y.V2 could be associated with a higher severity of respiratory failure at admission and a higher risk for developing VAP. We hypothesized that the expected gain in survival brought by recent improvements in critical care management could have been mitigated by increased transmissibility of the new lineage leading to admission of more severe patients. The immunological role of VOC 20H/501Y.V2 in the propensity for VAP requires further investigations.

Automated location of steel truss bridge damage using machine learning and raw strain sensor data

Strategic major infrastructure ageing requires structural health monitoring usage to avoid critical safety issues and disasters. Machine Learning can be a valuable tool to automate the process of analysing raw monitoring data. Usually, frequency domain damage-sensitive features are extracted with data pre-processing procedures; thus these features are used as input for classification or regression problems.This paper describes a method of locating damage in steel truss railway bridges through machine learning classification tools, enabling automatic analysis of raw strain sensors signals without any pre-processing or preliminary feature extraction. Data were generated by simulating different damage scenarios with a finite element software, and then were processed by two machine learning classification tools: (a) the K-nearest Neighbours was adopted with the Dynamic Time Warping algorithm metric to select the most informative features; (b) a model suitable for high-dimensional data analysis, known as the Convolutional Neural Network, was then trained to classify strain sensors time series. The results indicate that the method applied can detect damages with an accuracy of 93% and is suitable for structural health monitoring.

Spatial Distribution of Aedes aegypti Oviposition Temporal Patterns and Their Relationship with Environment and Dengue Incidence.

Simple SummaryAedes aegypti, the mosquito species that transmits dengue virus among others, is fully adapted to thrive in urban areas. Their activity, however, varies in time and space and this might imply different transmission risk. We hypothesize that the temporal differences in mosquito activity are determined by local environmental conditions. Hence, we explore the existence of groups of temporal patterns in weekly time series of ovitraps records and we associate those patterns to environmental variables derived from remote sensing data and also to dengue incidence. We found three groups of temporal patterns that showed association with land cover diversity, heterogeneity and variability in vegetation and humidity indices estimated over 50-m radius buffer areas surrounding ovitraps. Dengue incidence on a neighborhood basis showed a weak but positive association with the percentage of pixels belonging to one of the patterns detected. The understanding of the spatial distribution of temporal patterns and their environmental determinants might then become relevant to guide the allocation of prevention and monitoring interventions.Aedes aegypti, the mosquito species transmitting dengue, zika, chikungunya and yellow fever viruses, is fully adapted to thrive in urban areas. The temporal activity of this mosquito, however, varies within urban areas which might imply different transmission risk. In this work, we hypothesize that temporal differences in mosquito activity patterns are determined by local environmental conditions. Hence, we explore the existence of groups of temporal patterns in weekly time series of Ae. aegypti ovitraps records (2017–2019) by means of time series clustering. Next, with the aim of predicting risk in places with no mosquito field data, we use machine learning classification tools to assess the association of temporal patterns with environmental variables derived from satellite imagery and predict temporal patterns over the city area to finally test the relationship with dengue incidence. We found three groups of temporal patterns that showed association with land cover diversity, variability in vegetation and humidity and, heterogeneity measured by texture indices estimated over buffer areas surrounding ovitraps. Dengue incidence on a neighborhood basis showed a weak but positive association with the percentage of pixels belonging to only one of the temporal patterns detected. The understanding of the spatial distribution of temporal patterns and their environmental determinants might then become highly relevant to guide the allocation of prevention and potential interventions. Further investigation is still needed though to incorporate other determinants not considered here.

Gesture Recognition Using Reflected Visible and Infrared Lightwave Signals

In this article, we demonstrate the ability to recognize hand gestures in a noncontact wireless fashion using only incoherent light signals reflected from a human subject. Fundamentally distinguished from radar, lidar, and camera-based sensing systems, this sensing modality uses only a low-cost light source (e.g., LED) and a sensor (e.g., photodetector). The lightwave-based gesture recognition system identifies different gestures from the variations in light intensity reflected from the subject's hand within a short (20-35 cm) range. As users perform different gestures, scattered light forms unique, statistically repeatable, time-domain signatures. These signatures can be learned by repeated sampling to obtain the training model against which unknown gesture signals are tested and categorized. These time-domain variations of the lightwave signals reflected from hand are denoised, standardized, and then classified by using machine learning classification tools such as $K$-nearest neighbors and support vector machine. Performance evaluations have been conducted with eight gestures, five subjects, different distances and lighting conditions, and visible and infrared light sources. The results demonstrate the best hand gesture recognition performance of infrared sensing at 20 cm with an average of 96% accuracy. The developed gesture recognition system is low-cost, effective, and noncontact technology for numerous human-computer interaction applications.

Environmental risk characterization and ecological process determination of bacterial antibiotic resistome in lake sediments

The increasing prevalence of antibiotic resistance genes (ARGs) in aquatic environments has attracted considerable concerns due to their potential threat to public health. For reducing environmental risk of ARGs, it is crucial to identify the pathogenic resistant bacteria, determine the driving forces governing the ARG community and apportion their sources, which is yet remained to explore. In this study, we developed a framework integrating high-throughput sequencing (HTS) analyses, null-model-based methods and machine-learning classification tool for understanding the environmental resistome risk and the ecological processes that control the ARG profile in aquatic sediments, and applied to two urban lakes (Lake Tai and Lake Baiyang) in China. The HTS-based metagenomic analyses revealed abundant and diverse resistome, mobilome and virulome in the two lakes, including some emerging ARGs such as mcr and carbapenemases types. Relatively, the diversities for ARGs, mobile genetic elements (MGEs) and virulence factor genes in Lake Baiyang were significantly higher than those in Lake Tai (p < 0.05). The metagenomic assembly and binning approaches tracked a number of potential pathogenic antibiotic resistant bacteria and found the co-occurrence of ARGs, MGEs and human bacterial pathogens in ~50% of the sediment samples, indicating a substantial resistome risk in the lakes. Comparison of multiple-site beta-diversity dissimilarity indexes suggested the ARG diversity was mainly explained by the spatial turnover rather than nestedness and exhibited significant distance-decay pattern. The results of using a novel null-model-based stochasticity ratio showed the stochastic processes made a higher contribution than the deterministic processes on the ARG profile in the environment, especially for Lake Baiyang (>65%). This was confirmed by the determination analyses of various ecological processes on ARG community by utilizing the null-model-based statistical framework for quantifying community assembly. That is, homogenizing dispersal (40%) dominated in Lake Baiyang, followed by homogeneous selection (32%) and ecological drift (15%), while ecological drift (33%) and homogenizing dispersal (31%) were the dominators in Lake Baiyang. SourceTracker analysis showed human sewage-associated sources were the largest contributor (~62%) of ARGs in the environment. The findings shed light on the dissemination risk and driver dynamics of antimicrobial resistance in the aquatic environment, which may help to make effective management strategies for controlling pollution of ARGs.

Pore system characterization of organic-rich shales using nanoscale-resolution 3D imaging

The Vaca Muerta Shale in Argentina is the first major commercial shale oil/gas play outside of North America. High-resolution 2D/3D imaging of shale rocks for the purpose of establishing their mineralogy, total, and connected porosities has become more and more sophisticated. In this paper, nanoscale-resolution focused ion beam (FIB)–scanning electron microscopy (SEM) nano-tomography was used to obtain images of pore structures within two organic-rich regions of interest (ROIs), selected based on correlative SEM and automated mineralogy maps. Advanced machine learning classification tools were used to segment the images and assign porosities and other components. Pore size distribution and pore connectivity analyzes revealed that about 95% of all the pores, present within the two ROIs, had a diameter of less than approximately 75 nm, and that most of these pores were poorly connected. In a similar fashion, the flow rate distribution analysis showed that pores with diameters of about 150–330 nm contributed to over 50% of the flow capacity of the connected pore systems. These results suggest that although most of the pores typically found in shales have pore diameter smaller than about 100 nm, most of the hydrocarbon production may be carried by a relatively small number of larger connected pores with pore diameter greater than about 150 nm. This study implies that a large portion of the organic-hosted pores (with diameter typically smaller than about 100 nm) may not provide permeable flow pathways for the oil and/or gas migration, and hence may have very little contribution to the hydrocarbon production.

Source identification of antibiotic resistance genes in a peri-urban river using novel crAssphage marker genes and metagenomic signatures

Antimicrobial resistance is a growing public health concern, and environment is regarded as an important reservoir and dissemination route for antibiotic resistance genes (ARGs). To prevent and control ARG pollution, it is essential to correctly disentangle source-sink relationship of ARGs in the environment. However, accurately apportioning sources of ARGs is still a big challenge due to the complex interaction of multiple sources and contaminants in the environment with changing dynamics. In this study, we addressed this problem and focused on identifying the potential sources of ARGs in a peri-urban river by jointly utilizing two novel microbial source tracking methods. To attain the objective, sediment/water samples were collected from the peri-urban river and four ARG-associated ecotypes including effluents of sewage treatment plants (STPs), STP influent, chicken manures and pig manures. The high-throughput profilings of ARGs and microbial taxa in the river sediments and the four ecotypes were comprehensively characterized in combination of shotgun sequencing and metagenomic assembly analysis. CrAssphage, a recently-discovered DNA bacteriophage, was employed to track the impact of human fecal pollution on ARGs in the river sediments. Further, SourceTracker, a machine-learning classification tool, was used for quantifying the contributions of potential sources to ARGs in the river sediments based on the metagenomic signatures of ARGs and microbial taxa. In total, 888 ARG subtypes belonging to 29 ARG types were detected across all samples, including mcr-1 and a range of carbapenemases types. Statistical analyses suggested different ecotypes generally had distinct profiles of both ARGs and microbial taxa, while the ARG compositions were significantly correlated with the microbial community. Source tracking with crAssphage showed the presence of ARGs in the river sediments might be largely impacted by the extent of human fecal pollution, which was also confirmed by the analyses of SourceTracker that the discharge from STPs was the largest contributor of ARGs (81.6–92.1%) and microbes (49.3–68.1%) in the river sediments. Results of the study can help us to better understand the characterization of ARGs in the peri-urban ecosystem and to design effective prevention and control strategies for reducing ARG dissemination.

Machine Learning models for the Identification of Cognitive Tasks using Autonomic Reactions from Heart Rate Variability and Electrodermal Activity.

Indices of heart rate variability (HRV) and electrodermal activity (EDA), in conjunction with machine learning models, were used to identify one of three tasks a subject is performing based on autonomic response elicited by the specific task. Using non-invasive measures to identify the task performed by a subject can help to provide individual monitoring and guidance to avoid the consequences of reduced performance due to fatigue or other stressors. In the present study, sixteen subjects were enrolled to undergo three tasks: The psychomotor vigilance task (PVT), an auditory working memory task (the n-back paradigm), and a visual search (ship search, SS). Electrocardiogram (ECG) (for HRV analysis) and EDA data were collected during the tests. For task-classification, we tested four machine learning classification tools: k-nearest neighbor classifier (KNN), support vector machines (SVM), decision trees, and discriminant analysis (DA). Leave-one-subject-out cross-validation was used to evaluate the performance of the constructed models to prevent overfitting. The most accurate models were the KNN (66%), linear SVM (62%), and linear DA (62%). The results of this study showed that it is possible to identify the task a subject is performing based on the subject’s autonomic reactions (from HRV and EDA). This information can be used to monitor individuals within a larger group to assist in reducing errors caused by uncoordinated or poor performance by allowing for automated tracking of and communication between individuals.

Sentiment Analysis of Arabic Jordanian Dialect Tweets

Sentiment Analysis (SA) of social media contents has become one of the growing areas of research in data mining. SA provides the ability of text mining the public opinions of a subjective manner in real time. This paper proposes a SA model of Arabic Jordanian dialect tweets. Tweets are annotated on three different classes; positive, negative, and neutral. Support Vector Machines (SVM) and Naive Bayes (NB) are used as supervised machine learning classification tools. Preprocessing of such tweets for SA is done via; cleaning noisy tweets, normalization, tokenization, namely, Entity Recognition, removing stop words, and stemming. The results of the experiments conducted on this model showed encouraging outcomes when Arabic light stemmer/segment is applied on Arabic Jordanian dialect tweets. Also, the results showed that SVM has better performance than NB on such tweets’ classifications.

Analysis of drug resistance in HIV protease

BackgroundDrug resistance in HIV is the major problem limiting effective antiviral therapy. Computational techniques for predicting drug resistance profiles from genomic data can accelerate the appropriate choice of therapy. These techniques can also be used to select protease mutants for experimental studies of resistance and thereby assist in the development of next-generation therapies.ResultsThe machine learning produced highly accurate and robust classification of HIV protease resistance. Genotype data were mapped to the enzyme structure and encoded using Delaunay triangulation. Generative machine learning models trained on one inhibitor could classify resistance from other inhibitors with varying levels of accuracy. Generally, the accuracy was best when the inhibitors were chemically similar.ConclusionsRestricted Boltzmann Machines are an effective machine learning tool for classification of genomic and structural data. They can also be used to compare resistance profiles of different protease inhibitors.

Causal data science for financial stress testing

The most recent financial upheavals have cast doubt on the adequacy of some of the conventional quantitative risk management strategies, such as VaR (Value at Risk), in many common situations. Consequently, there has been an increasing need for verisimilar financial stress testings, namely simulating and analyzing financial portfolios in extreme, albeit rare scenarios. Unlike conventional risk management which exploits statistical correlations among financial instruments, here we focus our analysis on the notion of probabilistic causation, which is embodied by Suppes-Bayes Causal Networks (SBCNs); SBCNs are probabilistic graphical models that have many attractive features in terms of more accurate causal analysis for generating financial stress scenarios.In this paper, we present a novel approach for conducting stress testing of financial portfolios based on SBCNs in combination with classical machine learning classification tools. The resulting method is shown to be capable of correctly discovering the causal relationships among financial factors that affect the portfolios and thus, simulating stress testing scenarios with a higher accuracy and lower computational complexity than conventional Monte Carlo simulations.

Forecasting of future stock prices using neural networks and genetic algorithms

Neural networks are a well established and widely used class of machine learning tools for classification and clustering that have been successfully applied to time-series analysis and prediction. On the other hand, genetic algorithms have been used in the literature for a vast range of optimisation problems ranging from electromagnetic optimisation to mechanical design, industrial control and genetic engineering. In this work, we propose to use the former in predicting future values of a time-series of particular interest, i.e., the future values of stock market indices. Based on a large body of work that is present in the literature, we develop, test and present a set of neural networks for predicting future stock market index values. Furthermore, we evaluate the use of modified GAs as a stand-alone tool for prediction, but also the use of GAs as neural network training and optimising tools. We also test two benchmark time-series extrapolation techniques based on linear regression. The proposed stock market prediction tools are fine-tuned and applied to a number of stock market index time-series and numerical results are presented demonstrating their superiority compared to standard benchmark techniques.

Efficient Simulation of Financial Stress Testing Scenarios with Suppes-Bayes Causal Networks

The most recent financial upheavals have cast doubt on the adequacy of some of the conventional quantitative risk management strategies, such as VaR (Value at Risk), in many common situations. Consequently, there has been an increasing need for verisimilar financial stress testings, namely simulating and analyzing financial portfolios in extreme, albeit rare scenarios. Unlike conventional risk management which exploits statistical correlations among financial instruments, here we focus our analysis on the notion of probabilistic causation, which is embodied by Suppes-Bayes Causal Networks (SBCNs), SBCNs are probabilistic graphical models that have many attractive features in terms of more accurate causal analysis for generating financial stress scenarios.In this paper, we present a novel approach for conducting stress testing of financial portfolios based on SBCNs in combination with classical machine learning classification tools. The resulting method is shown to be capable of correctly discovering the causal relationships among financial factors that affect the portfolios and thus, simulating stress testing scenarios with a higher accuracy and lower computational complexity than conventional Monte Carlo Simulations.

An alternative approach to the determination of scaling law expressions for the L–H transition in Tokamaks utilizing classification tools instead of regression

A new approach to determine the power law expressions for the threshold between the H and L mode of confinement is presented. The method is based on two powerful machine learning tools for classification: neural networks and support vector machines. Using as inputs clear examples of the systems on either side of the transition, the machine learning tools learn the input–output mapping corresponding to the equations of the boundary separating the confinement regimes. Systematic tests with synthetic data show that the machine learning tools provide results competitive with traditional statistical regression and more robust against random noise and systematic errors. The developed tools have then been applied to the multi-machine International Tokamak Physics Activity International Global Threshold Database of validated ITER-like Tokamak discharges. The machine learning tools converge on the same scaling law parameters obtained with non-linear regression. On the other hand, the developed tools allow a reduction of 50% of the uncertainty in the extrapolations to ITER. Therefore the proposed approach can effectively complement traditional regression since its application poses much less stringent requirements on the experimental data, to be used to determine the scaling laws, because they do not require examples exactly at the moment of the transition.

Codebook-Based Audio Feature Representation for Music Information Retrieval

Digital music has become prolific in the web in recent decades. Automated recommendation systems are essential for users to discover music they love and for artists to reach appropriate audience. When manual annotations and user preference data is lacking (e.g. for new artists) these systems must rely on content based methods. Besides powerful machine learning tools for classification and retrieval, a key component for successful recommendation is the audio content representation. Good representations should capture informative musical patterns in the audio signal of songs. These representations should be concise, to enable efficient (low storage, easy indexing, fast search) management of huge music repositories, and should also be easy and fast to compute, to enable real-time interaction with a user supplying new songs to the system. Before designing new audio features, we explore the usage of traditional local features, while adding a stage of encoding with a pre-computed codebook and a stage of pooling to get compact vectorial representations. We experiment with different encoding methods, namely the LASSO, vector quantization (VQ) and cosine similarity (CS). We evaluate the representations' quality in two music information retrieval applications: query-by-tag and query-by-example. Our results show that concise representations can be used for successful performance in both applications. We recommend using top- τ VQ encoding, which consistently performs well in both applications, and requires much less computation time than the LASSO.

Self-adaptive attribute weighting for Naive Bayes classification

Naive Bayes (NB) is a popular machine learning tool for classification, due to its simplicity, high computational efficiency, and good classification accuracy, especially for high dimensional data such as texts. In reality, the pronounced advantage of NB is often challenged by the strong conditional independence assumption between attributes, which may deteriorate the classification performance. Accordingly, numerous efforts have been made to improve NB, by using approaches such as structure extension, attribute selection, attribute weighting, instance weighting, local learning and so on. In this paper, we propose a new Artificial Immune System (AIS) based self-adaptive attribute weighting method for Naive Bayes classification. The proposed method, namely AISWNB, uses immunity theory in Artificial Immune Systems to search optimal attribute weight values, where self-adjusted weight values will alleviate the conditional independence assumption and help calculate the conditional probability in an accurate way. One noticeable advantage of AISWNB is that the unique immune system based evolutionary computation process, including initialization, clone, section, and mutation, ensures that AISWNB can adjust itself to the data without explicit specification of functional or distributional forms of the underlying model. As a result, AISWNB can obtain good attribute weight values during the learning process. Experiments and comparisons on 36 machine learning benchmark data sets and six image classification data sets demonstrate that AISWNB significantly outperforms its peers in classification accuracy, class probability estimation, and class ranking performance.