Continuous Attributes Research Articles

A Crowdsourcing Framework for Collecting Tabular Data

In crowdsourcing, human workers are employed to tackle problems that are traditionally difficult for computers (e.g., data cleaning, missing value filling, and sentiment analysis). In this paper, we study the effective use of crowdsourcing in filling missing values in a given relation (e.g., a table containing different attributes of celebrity stars, such as nationality and age). A task given to a worker typically consists of questions about the missing attribute values (e.g., What is the age of Jet Li?). Although this problem has been studied before, existing work often treats related attributes independently, leading to suboptimal performance. In this paper, we present T-Crowd, which is a crowdsourcing system that considers attribute relationships. Particularly, T-Crowd integrates each worker's answers on different attributes to effectively learn his/her trustworthiness and the true data values. The attribute relationship information is used to guide task allocation to workers. Our solution seamlessly supports categorical and continuous attributes. Our extensive experiments on real and synthetic datasets show that T-Crowd outperforms state-of-the-art methods, improving the quality of truth inference and reducing the monetary cost of crowdsourcing.

Extracting Classification Rules from Artificial Neural Network Trained with Discretized Inputs

Rule extraction from artificial neural networks remains important task in complex diseases such as diabetes and breast cancer where the rules should be accurate and comprehensible. The quality of rules is improved by the improvement of the network classification accuracy which is done by the discretization of input attributes. In this paper, we developed a rule extraction algorithm based on multiobjective genetic algorithms and association rules mining to extract highly accurate and comprehensible classification rules from ANN’s that have been trained using the discretization of the continuous attributes. The data pre-processing provides very good improvement of the ANN accuracy and consequently leads to improve the performance of the classification rules in terms of fidelity and coverage. The results show that our algorithm is very suitable for medical decision making, so an excellent average accuracy of 94.73 has been achieved for the Pima dataset and 99.36 for the breast cancer dataset.

An ontology enabled internet of things framework in intelligent agriculture for preventing post-harvest losses

Constituting the agriculture solid substance manufacture, the post-harvest sector processing schema is direct to preventing reduce the losses in intelligent agriculture. Many processing schemata will be preventing post-harvest losses on the agriculture solid substance manufacture, especially sekai-ichi apple is the regularly used fruit also used to make active in human-related activities of the sensory and control function consisting of an agricultural industry. Sekai-ichi apple is being a definite number of diseases induce, but it is to the highest degree of wastage involving in the Post-Harvest process. Especially sekai-ichi apple count loss is an unsafe many time because it not critically post-harvest. Regardless of consideration, the existing hierarchical model specified post-harvest losses prevention research has deficiencies to precise and quick detection of wastage for ensuring healthy separation of agriculture surroundings. This paper suggests a “Hierarchical Model within Ontology Enabled IoT” for distinguishable healthy separation of sekai-ichi apple by using Boosted Continuous Non-spatial whole Attribute Extraction (BCNAE). Sekai-ichi apple count loss is always safe on critically post-harvest. Proposed Post-Harvest hierarchical model specified post-harvest losses prevention and deficiencies to precise and quick detection of wastage for ensuring healthy separation of agriculture surroundings. In these suggestions, the separation cognitive operation takes the three levels of processing schemes such as lower level, middle level, and higher level. Firstly, the lower level is express agreements with the dynamic functioning for maintaining the definite number of manual induces. This lower level showing an absorption with the activity of manual separation by the human reliability determination. Secondly, the middle level is an express arrangement with the dynamic functioning for reducing the overfitting and accommodate to fitting the right shape deliberation. Middle level is establishing being generalized by concentrating the time-varying features in the occurrence of a change for the worse identification. Finally, the upper level is express for features refining with the help of the function of sekai-ichi apple image segmentation connection. This interpretability process helps to make the proven position of a prominent classification in a particular fruit on the agriculture solid substance. These three processing flow constructs the ontology structure with manually collected sekai-ichi apple images from a 3D sensor. The observational consequences express that the proposed BCNAE framework recognizes a detection performance carrying out with an optimized—separation ratio for time-variant of the separation process.

Modeling and verification of running process control for underground mine locomotive

The traditional underground mine locomotive unmanned system is difficult to define requirements accurately and guarantee the functional safety and integrity. To solve this problem, a solution based on Event-B formal method is put forward. In this paper, we select the locomotive running process control for formal modeling and verification, which is the core of the underground mine locomotive unmanned system. Based on the Event-B formal method and Rodin platform, five refinements are proposed to model and verify the locomotive running process control with both discrete and continuous attributes. The correctness, consistency and deadlock-freeness of the system design in the actual situation are verified with extensive simulations. The verification and simulation results show that the solution meets the design requirements of the running process control of the underground mine locomotive.

Discovering latent activity patterns from transit smart card data: A spatiotemporal topic model

Although automatically collected human travel records can accurately capture the time and location of human movements, they do not directly explain the hidden semantic structures behind the data, e.g., activity types. This work proposes a probabilistic topic model, adapted from Latent Dirichlet Allocation (LDA), to discover representative and interpretable activity categorization from individual-level spatiotemporal data in an unsupervised manner. Specifically, the activity-travel episodes of an individual user are treated as words in a document, and each topic is a distribution over space and time that corresponds to certain type of activity. The model accounts for a mixture of discrete and continuous attributes—the location, start time of day, start day of week, and duration of each activity episode. The proposed methodology is demonstrated using pseudonymized transit smart card data from London, U.K. The results show that the model can successfully distinguish the three most basic types of activities—home, work, and other. As the specified number of activity categories increases, more specific subpatterns for home and work emerge, and both the goodness of fit and predictive performance for travel behavior improve. This work makes it possible to enrich human mobility data with representative and interpretable activity patterns without relying on predefined activity categories or heuristic rules.

Holistic Affect Recognition Using PaNDA: Paralinguistic Non-Metric Dimensional Analysis

Humans perceive emotion from each other using a holistic perspective, accounting for diverse personal, non-emotional variables, such as age and personality, that shape expression. In contrast, today’s algorithms are mainly designed to recognize emotion in isolation, and are usually demonstrated only within one relatively narrow database. In this article, we propose a multi-task learning approach to jointly learn the recognition of affective states from speech along with various speaker attributes. A problem with multi-task learning is that sometimes inductive transfer can negatively impact performance. To mitigate negative transfer, we introduce the Paralinguistic Non-metric Dimensional Analysis (PaNDA) method that systematically measures task relatedness and also enables visualizing the topology of affective phenomena as a whole. In addition, we present a generic framework that conflates the concepts of single-task and multi-task learning. Using this framework, we construct two models that demonstrate holistic affect recognition: one treats all tasks as equally related, whereas the other one incorporates the task correlations between a main task and its supporting tasks obtained from PaNDA. Both models employ a multi-task deep neural network, in which separate output layers are used to predict discrete and continuous attributes, while hidden layers are shared across different tasks. On average across 18 classification and regression tasks, the weighted multi-task learning with PaNDA significantly improves performance compared to single-task and unweighted multi-task learning.

Speech emotion recognition using fusion of three multi-task learning-based classifiers: HSF-DNN, MS-CNN and LLD-RNN

Speech emotion recognition plays an increasingly important role in emotional computing and is still a challenging task due to its complexity. In this study, we developed a framework integrating three distinctive classifiers: a deep neural network (DNN), a convolution neural network (CNN), and a recurrent neural network (RNN). The framework was used for categorical recognition of four discrete emotions (i.e., angry, happy, neutral and sad). Frame-level low-level descriptors (LLDs), segment-level mel-spectrograms (MS), and utterance-level outputs of high-level statistical functions (HSFs) on LLDs were passed to RNN, CNN, and DNN, separately. Three individual models of LLD-RNN, MS-CNN, and HSF-DNN were obtained. In the models of MS-CNN and LLD-RNN, the attention mechanism based weighted-pooling method was utilized to aggregate the CNN and RNN outputs. To effectively utilize the interdependencies between the two approaches of emotion description (discrete emotion categories and continuous emotion attributes), a multi-task learning strategy was implemented in these three models to acquire generalized features by simultaneously operating classification of discrete categories and regression of continuous attributes. Finally, a confidence-based fusion strategy was developed to integrate the power of different classifiers in recognizing different emotional states. Three experiments on emotion recognition based on the IEMOCAP corpus were conducted. Our experimental results show that the weighted pooling method based on attention mechanism endowed the neural networks with the capability to focus on emotionally salient parts. The generalized features learned in the multi-task learning helped the neural networks to achieve higher accuracies in the tasks of emotion classification. Furthermore, our proposed fusion system achieved weighted accuracy of 57.1% and unweighted accuracy of 58.3%, which were significantly higher than those of each individual classifier. The effectiveness of the proposed approach based on classifier fusion was thus validated.

Novel Grey Relational Feature Extraction Algorithm for Software Fault-Proneness Using BBO (B-GRA)

The inherent uncertainty of software gives a vague and imprecise solution when it is solved by human judgment. As the project expands, the issues of missing data values, outlier detection, feature subset selection and prediction of faultiness behaviour should be addressed. The feature selection process may lead to the production of high-dimensional data sets that may contribute to many irrelevant or redundant features. In this paper, we focussed on the optimal feature subset selection and fault prediction at the early stage of a project. We propose the novel approach of grey relational analysis (GRA) from grey system theory by optimizing the grey relational grade function using biogeography optimization referred to as B-GRA. The proposed algorithm gives resilience to users to select features for both continuous and categorical attributes. The issues such as feature subset selection, heterogeneity of data sets, outlier analysis and fault prediction are addressed, and then, B-GRA and GRA approaches on five publically available data sets are evaluated using statistical and machine learning techniques. Experimental results show significant results indicating that the proposed methodology can be used for the prediction of faults and produce conceivable results when compared with the GRA feature selection approach.

FMDBN: A first-order Markov dynamic Bayesian network classifier with continuous attributes

With the development of data driven decision making and prediction, time-series data are ubiquitous and the demand for its classification is vast. Although a large body of research has been reported in the literature, it is mainly oriented to situations in which class and attributes are changing simultaneously. In practice however, those class and attributes changes are not always synchronous. This means that further studies for asynchronous classifier problems are necessary. In this paper, a first-order Markov dynamic Bayesian network classifier is proposed to address the asynchronous issue, by combing time-series data preprocessing, time-delayed and dislocated transformation of variables, initial and evolutionary learning. The attribute density in this classifier is estimated based on Gaussian function, and the classification accuracy criterion for time-series progressiveness is also considered. This classifier has a relatively simple structure, which can avoid the problem of overfitting. In addition, data can effectively be classified by utilizing three kinds of classification information, namely time-delayed, non-time-delayed and mixed information in multivariate time-series datasets. The proposed method is also able to accumulate classification information via iterative evolution and thus improve the generalization of classifiers. Experiments were carried out by using standard time-series datasets from UCI, financial and macroeconomic domains. The experimental results show that the proposed first-order Markov dynamic Bayesian network classifier is more accurate in dealing with these dynamic classification problems.

Optimizing the C4.5 Decision Tree Algorithm using MSD-Splitting

We propose an optimization of Dr. Ross Quin-lan’s C4.5 decision tree algorithm, used for data mining and classification. We will show that by discretizing and binning a data set’s continuous attributes into four groups using our novel technique called MSD-Splitting, we can significantly improve both the algorithm’s accuracy and efficiency, especially when applied to large data sets. We applied both the standard C4.5 algorithm and our optimized C4.5 algorithm to two data sets obtained from UC Irvine’s Machine Learning Repository: Census Income and Heart Disease. In our initial model, we discretized continuous attributes by splitting them into two groups at the point with the minimum expected information requirement, in accordance with the standard C4.5 algorithm. Using five-fold cross-validation, we calculated the average accuracy of our initial model for each data set. Our initial model yielded a 75.72% average accuracy across both data sets. The average execution time of our initial model was 1,541.57 s for the Census Income data set and 50.54 s for the Heart Disease data set. We then optimized our model by applying MSD-Splitting, which discretizes continuous attributes by splitting them into four groups using the mean and the two values one standard deviation away from the mean as split points. The accuracy of our model improved by an average of 5.11%across both data sets, while the average execution time reduced by an average of 96.72% for the larger Census Income data set and 46.38% for the Heart Disease data set.

MC4.5 decision tree algorithm: an improved use of continuous attributes

C4.5 is one of the top ten data mining algorithms; it is the most widely used decision trees construction techniques. Although effective, it suffers from the problem of complexity when it deals with continuous attributes. It also leads to a certain level of information loss. Therefore, minimising such loss and reducing the time complexity is one of the main goals in this paper. With the intention of alleviating these problems, this paper presents a novel algorithm namely MC4.5, which proposes the statistical mean as an alternative to the C4.5 threshold selection process. To demonstrate the effectiveness of the new algorithm, a complete evaluation was launched to prove that MC4.5 complies with the objectives previously mentioned. From the theoretical perspective, we develop an analysis of the complexity to compare algorithms. Empirically, we conduct an experimental study using 30 datasets to prove that, in most cases, the proposed algorithm leads to smaller decision trees with better accuracy comparing to the C4.5 algorithm.

An Improved C4.5 Algorthm in Bagging Integration Model

The C4.5 algorithm has three shortcomings: the wide range of candidate segmentation threshold sequences for continuous attributes, the comprehensive influence of different attributes and local subsets under the same attribute, and the inter-attribute redundancy. When dealing with continuous attributes, sampling and threshold supplement processing near the transition boundary of the attribute interval corresponding to the adjacent different categories are performed for narrowing the range of candate segmentation threshold sequences. By adding standardizing Euclidean distance of the attribute global and local factors to represent attribute weight, the calculation of C4.5 information gain is otpimized. And averaging Gini index of other attributes and adding correction factor, the influence of redundancy between attributes is greatly decreased. The overall average improvement range of the base classifier and the bagging integration classifier is 0.6%~2.1% and 0.7% ~ 2.7%, respectively, which shows that this integration model can improve the classification accuracy and also validate its feasibility and reliability.

Grey relational classification algorithm for software fault proneness with SOM clustering

The estimation by the human judgment to deal with the inherent uncertainty of software gives a vague and imprecise solution. To cope with this challenge, we propose a new hybrid analogy model based on the integration of grey relational analysis (GRA) classification with self-organising map (SOM) clustering. In this paper, a new classification approach is proposed to distribute the data to similar groups. The attributes are selected based on GRC values. In the proposed, the similarity measure between reference project and cluster head is computed to determine the cluster to which target project belongs. The fault-proneness of reference project is estimated based on the regression equation of the selected cluster. The proposed algorithm gives resilience to users to select features for both continuous and categorical attributes. In this study, two scenarios based on the integration of proposed classification with regression have been proposed. Experimental results show significant results indicating that proposed methodology can be used for the prediction of faults and produce conceivable results when compared with the results of multilayer-perceptron, logistic regression, bagging, naive Bayes and sequential minimal optimisation (SMO).

Research on an Ensemble Classification Algorithm Based on Differential Privacy

In the field of information security, privacy protection based on machine learning is currently a hot topic. Combining differential privacy protection with AdaBoost, a machine learning ensemble classification algorithm, this paper proposes a scheme under differential privacy named CART-DPsAdaBoost (CART-Differential privacy structure of AdaBoost). In the process of boosting, the algorithm combines the idea of bagging, and uses a classification and regression tree (CART) stump as the base learner for ensemble learning. Applying feature perturbation, based on a random subspace algorithm, the exponential mechanism is used to select the splitting point for continuous attributes. We use the Gini index to find the optimal binary partitioning point for discrete attributes and add noise according to the Laplace mechanism. Throughout the process, a privacy budget is allocated in order to meet the appropriate differential privacy protection needs for the current application. Unlike similar algorithms, this method does not require discretization during preprocessing of the data. Experimental results with the Census Income, Digit Recognizer, and Adult Data Set show that while protecting private information, the scheme has little impact on classification accuracy and can effectively address large-scale and high-dimensional data classification problems.

Discretization Using Combination of Heuristics for High Accuracy with Huge Noise Reduction

Over the years, several algorithms for discretization have been devised, but the problem of efficient, accurate discretization still remains an open problem. This paper proposes a novel discretization algorithm, called SPID5, based on combination of two heuristics, one being local and the other global, both being supervised and their combination resulting in a significant synergy. The local heuristic is the well-known information gain of the continuous attributes, and the global heuristic is a novel concept of iterative reduction of noise in the data set. The reduction of noise is achieved by reducing successively pseudo deletion count of the data set to be discretized. The performance of SPID5 algorithm is compared with that of three well-known and time-tested discretization algorithms, using six state-of-the-art classifiers and 35 real-world data sets from the standard UCI data repository. Performance of SPID5 compares favorably with that of all the three existing discretization algorithms it is compared with, not only in terms of classification accuracy but also in terms of noise reduction in the data sets.

MC4.5 decision tree algorithm: an improved use of continuous attributes

C4.5 is one of the top ten data mining algorithms; it is the most widely used decision trees construction techniques. Although effective, it suffers from the problem of complexity when it deals wit...

An Adaptive Hybrid Ant Colony Optimization Algorithm for The Classification Problem

Classification is an important data analysis and data mining technique. Taking into account the comprehensibility of the classifier generated, an adaptive hybrid ant colony optimization algorithm called A_HACO is proposed which can effectively solve classification problem and get the comprehensible classification rules at the same time. The algorithm incorporates the artificial bee colony optimization strategy into the ant colony algorithm. The ant colony global optimization process is used to adaptively select the appropriate rule evaluation function for the data set given. Based on the classification rules obtained, the artificial bee colony optimization strategy is used to tackle the continuous attributes for further optimization of classification rules. This approach is evaluated experimentally using different standard real datasets, and compared with some proposed related classification algorithms. It shows that A_HACO can adaptively select the appropriate rule evaluation function and has better accuracy compared with related works.

Supervised non-parametric discretization based on Kernel density estimation

Nowadays, machine learning algorithms can be found in many applications where the classifiers play a key role. In this context, discretizing continuous attributes is a common step previous to classification tasks, the main goal being to retain as much discriminative information as possible. In this paper, we propose a supervised univariate non-parametric discretization algorithm which allows the use of a given supervised score criterion for selecting the best cut points. The candidate cut points are evaluated by computing the selected score value using kernel density estimation. The computational complexity of the proposed procedure is O(NlogN), where N is the length of the data. Our proposed algorithm generates a low complexity in discretization policies while retaining the discriminative information of the original continuous variables. In order to assess the validity of the proposed method, a set of real and artificial datasets has been used and the results show that the algorithm provides competitive results in terms of performance, a low complexity in the discretization policies and a high performance.

Improved Classification Techniques to Predict the Co-disease in Diabetic Mellitus Patients using Discretization and Apriori Algorithm

The demand for data mining is now unavoidable in the medical industry due to its various applications and uses in predicting the diseases at the early stage. The methods available in the data mining theories are easy to extract the useful patterns and speed to recognize the task based outcomes. In data mining the classification models are really useful in building the classes for the medical data sets for future analysis in an accurate way. Besides these facilities, Association rules in data mining are a promising technique to find hidden patterns in a medical data set and have been successfully applied with market basket data, census data and financial data. Apriori algorithm, is considered to be a classic algorithm, is useful in mining frequent item sets on a database containing a large number of transactions and it also predicts the relevant association rules. Association rules capture the relationship of items that are present in data sets and when the data set contains continuous attributes, the existing algorithms may not work due to this, discretization can be applied to the association rules in order to find the relation between various patterns in data set. In this paper of our research, using Discretized Apriori the research work is done to predict the by-disease in people who are found with diabetic syndrome; also the rules extracted are analyzed. In the discretization step, numerical data is discretized and fed to the Apriori algorithm for better association rules to predict the diseases.

Detection of Lower Albuminuria Levels and Early Development of Diabetic Kidney Disease Using an Artificial Intelligence-Based Rule Extraction Approach

The aim of the present study was to determine the lowest cut-off value for albuminuria levels, which can be used to detect diabetic kidney disease (DKD) using the urinary albumin-to-creatinine ratio (UACR). National Health and Nutrition Examination Survey (NHANES) data for US adults were used, and participants were classified as having diabetes or prediabetes based on a self-report and physiological measures. The study dataset comprised 942 diabetes and 524 prediabetes samples. This study clarified the significance of the lower albuminuria (UACR) levels, which can detect DKD, using an artificial intelligence-based rule extraction approach. The diagnostic rules (15 concrete rules) for both samples were extracted using a recursive-rule eXtraction (Re-RX) algorithm with continuous attributes (continuous Re-RX) to discriminate between prediabetes and diabetes datasets. Continuous Re-RX showed high test accuracy (77.56%) and a large area under the receiver operating characteristics curve (75%), which derived the two cut-off values (6.1 mg/g Cr and 71.00 mg/g Cr) for the lower albuminuria level in the UACR to detect early development of DKD. The early cut-off values for normoalbuminuria (NA) and microalbuminuria (MA) will be determined to help detect CKD and DKD, and to detect diabetes before MA develop and to prevent diabetic complications.