Genetic Algorithms for Feature Generation in the Context of Audio Classification

Choosing good features is an essential part of machine learning. Recent techniques aim to automate this process. For instance, feature learning intends to learn the transformation of raw data into a useful representation to machine learning tasks. In automatic audio classification tasks, this is interesting since the audio, usually complex information, needs to be transformed into a computationally convenient input to process. Another technique tries to generate features by searching a feature space. Genetic algorithms, for instance, have being used to generate audio features by combining or modifying them. We find this approach particularly interesting and, despite the undeniable advances of feature learning approaches, we wanted to take a step forward in the use of genetic algorithms to find audio features, combining them with more conventional methods, like PCA, and inserting search control mechanisms, such as constraints over a confusion matrix. This work presents the results obtained on particular audio classification problems.

Comparison of Different k-NN Models for Speed Prediction in an Urban Traffic Network

A database that records average traffic speeds measured at five-minute intervals for all the links in the traffic network of a metropolitan city. While learning from this data the models that can predict future traffic speed would be beneficial for the applications such as the car navigation system, building predictive models for every link becomes a nontrivial job if the number of links in a given network is huge. An advantage of adopting k-nearest neighbor (k-NN) as predictive models is that it does not require any explicit model building. Instead, k-NN takes a long time to make a prediction because it needs to search for the k-nearest neighbors in the database at prediction time. In this paper, we investigate how much we can speed up k-NN in making traffic speed predictions by reducing the amount of data to be searched for without a significant sacrifice of prediction accuracy. The rationale behind this is that we had a better look at only the recent data because the traffic patterns not only repeat daily or weekly but also change over time. In our experiments, we build several different k-NN models employing different sets of features which are the current and past traffic speeds of the target link and the neighbor links in its up/down-stream. The performances of these models are compared by measuring the average prediction accuracy and the average time taken to make a prediction using various amounts of data.

Building a Scalable Telemetry Based Multiclass Predictive Maintenance Model in R

Many organizations are faced with the challenge of how to analyze and build Machine Learning models using their sensitive telemetry data. In this paper, we discuss how users can leverage the power of R without having to move their big data around as well as a cloud based solution for organizations willing to host their data in the cloud. By using ScaleR technology to benefit from parallelization and remote computing or R Services on premise or in the cloud, users can leverage the power of R at scale without having to move their data around.

Development of Prediction Models of Day-Ahead Hourly Building Electricity Consumption and Peak Power Demand Using the Machine Learning Method

To encourage building owners to purchase electricity at the wholesale market and reduce building peak demand, this study aims to develop models that predict day-ahead hourly electricity consumption and demand using artificial neural network (ANN) and support vector machine (SVM). All prediction models are built in Python, with tool Scikit-learn and Pybrain. The input data for both consumption and demand prediction are time stamp, outdoor dry bulb temperature, relative humidity, air handling unit (AHU), supply air temperature and solar radiation. Solar radiation, which is unavailable a day-ahead, is predicted at first, and then this estimation is used as an input to predict consumption and demand. Models to predict consumption and demand are trained in both SVM and ANN, and depend on cooling or heating, weekdays or weekends. The results show that ANN is the better option for both consumption and demand prediction. It can achieve 15.50% to 20.03% coefficient of variance of root mean square error (CVRMSE) for consumption prediction and 22.89% to 32.42% CVRMSE for demand prediction, respectively. To conclude, the presented models have potential to help building owners to purchase electricity at the wholesale market, but they are not robust when used in demand response control.

A Minimum Spanning Tree-Based Method for Initializing the K-Means Clustering Algorithm

The traditional k-means algorithm has been widely used as a simple and efficient clustering method. However, the algorithm often converges to local minima for the reason that it is sensitive to the initial cluster centers. In this paper, an algorithm for selecting initial cluster centers on the basis of minimum spanning tree (MST) is presented. The set of vertices in MST with same degree are regarded as a whole which is used to find the skeleton data points. Furthermore, a distance measure between the skeleton data points with consideration of degree and Euclidean distance is presented. Finally, MST-based initialization method for the k-means algorithm is presented, and the corresponding time complexity is analyzed as well. The presented algorithm is tested on five data sets from the UCI Machine Learning Repository. The experimental results illustrate the effectiveness of the presented algorithm compared to three existing initialization methods.

Context Detection in Spreadsheets Based on Automatically Inferred Table Schema

Programming requires years of training. With natural language and end user development methods, programming could become available to everyone. It enables end users to program their own devices and extend the functionality of the existing system without any knowledge of programming languages. In this paper, we describe an Interactive Spreadsheet Processing Module (ISPM), a natural language interface to spreadsheets that allows users to address ranges within the spreadsheet based on inferred table schema. Using the ISPM, end users are able to search for values in the schema of the table and to address the data in spreadsheets implicitly. Furthermore, it enables them to select and sort the spreadsheet data by using natural language. ISPM uses a machine learning technique to automatically infer areas within a spreadsheet, including different kinds of headers and data ranges. Since ranges can be identified from natural language queries, the end users can query the data using natural language. During the evaluation 12 undergraduate students were asked to perform operations (sum, sort, group and select) using the system and also Excel without ISPM interface, and the time taken for task completion was compared across the two systems. Only for the selection task did users take less time in Excel (since they directly selected the cells using the mouse) than in ISPM, by using natural language for end user software engineering, to overcome the present bottleneck of professional developers.

Chemical Reaction Algorithm for Expectation Maximization Clustering

Clustering is an intensive research for some years because of its multifaceted applications, such as biology, information retrieval, medicine, business and so on. The expectation maximization (EM) is a kind of algorithm framework in clustering methods, one of the ten algorithms of machine learning. Traditionally, optimization of objective function has been the standard approach in EM. Hence, research has investigated the utility of evolutionary computing and related techniques in the regard. Chemical Reaction Optimization (CRO) is a recently established method. So the property embedded in CRO is used to solve optimization problems. This paper presents an algorithm framework (EM-CRO) with modified CRO operators based on EM cluster problems. The hybrid algorithm is mainly to solve the problem of initial value sensitivity of the objective function optimization clustering algorithm. Our experiments mainly take the EM classic algorithm:k-means and fuzzy k-means as an example, through the CRO algorithm to optimize its initial value, get K-means-CRO and FKM-CRO algorithm. The experimental results of them show that there is improved efficiency for solving objective function optimization clustering problems.

Hand Gesture Interpretation Using Sensing Glove Integrated with Machine Learning Algorithms

In this paper, we present a low cost design for a smart glove that can perform sign language recognition to assist the speech impaired people. Specifically, we have designed and developed an Assistive Hand Gesture Interpreter that recognizes hand movements relevant to the American Sign Language (ASL) and translates them into text for display on a Thin-Film-Transistor Liquid Crystal Display (TFT LCD) screen as well as synthetic speech. Linear Bayes Classifiers and Multilayer Neural Networks have been used to classify 11 feature vectors obtained from the sensors on the glove into one of the 27 ASL alphabets and a predefined gesture for space. Three types of features are used; bending using six bend sensors, orientation in three dimensions using accelerometers and contacts at vital points using contact sensors. To gauge the performance of the presented design, the training database was prepared using five volunteers. The accuracy of the current version on the prepared dataset was found to be up to 99.3% for target user. The solution combines electronics, e-textile technology, sensor technology, embedded system and machine learning techniques to build a low cost wearable glove that is scrupulous, elegant and portable.

A Static Android Malware Detection Based on Actual Used Permissions Combination and API Calls

Android operating system has been recognized by most application developers because of its good open-source and compatibility, which enriches the categories of applications greatly. However, it has become the target of malware attackers due to the lack of strict security supervision mechanisms, which leads to the rapid growth of malware, thus bringing serious safety hazards to users. Therefore, it is critical to detect Android malware effectively. Generally, the permissions declared in the AndroidManifest.xml can reflect the function and behavior of the application to a large extent. Since current Android system has not any restrictions to the number of permissions that an application can request, developers tend to apply more than actually needed permissions in order to ensure the successful running of the application, which results in the abuse of permissions. However, some traditional detection methods only consider the requested permissions and ignore whether it is actually used, which leads to incorrect identification of some malwares. Therefore, a machine learning detection method based on the actually used permissions combination and API calls was put forward in this paper. Meanwhile, several experiments are conducted to evaluate our methodology. The result shows that it can detect unknown malware effectively with higher true positive rate and accuracy while maintaining a low false positive rate. Consequently, the AdaboostM1 (J48) classification algorithm based on information gain feature selection algorithm has the best detection result, which can achieve an accuracy of 99.8%, a true positive rate of 99.6% and a lowest false positive rate of 0.

Performance Comparison of Different Regression Methods for a Polymerization Process with Adaptive Sampling

Developing complete mechanistic models for polymerization reactors is not easy, because complex reactions occur simultaneously; there is a large number of kinetic parameters involved and sometimes the chemical and physical phenomena for mixtures involving polymers are poorly understood. To overcome these difficulties, empirical models based on sampled data can be used instead, namely regression methods typical of machine learning field. They have the ability to learn the trends of a process without any knowledge about its particular physical and chemical laws. Therefore, they are useful for modeling complex processes, such as the free radical polymerization of methyl methacrylate achieved in a batch bulk process. The goal is to generate accurate predictions of monomer conversion, numerical average molecular weight and gravimetrical average molecular weight. This process is associated with non-linear gel and glass effects. For this purpose, an adaptive sampling technique is presented, which can select more samples around the regions where the values have a higher variation. Several machine learning methods are used for the modeling and their performance is compared: support vector machines, k-nearest neighbor, k-nearest neighbor and random forest, as well as an original algorithm, large margin nearest neighbor regression. The suggested method provides very good results compared to the other well-known regression algorithms.

A Case-Based Reasoning-Decision Tree Hybrid System for Stock Selection

Stock selection is an important decision-making problem. Many machine learning and data mining technologies are employed to build automatic stock-selection system. A profitable stock-selection system should consider the stock’s investment value and the market timing. In this paper, we present a hybrid system including both engage for stock selection. This system uses a case-based reasoning (CBR) model to execute the stock classification, uses a decision-tree model to help with market timing and stock selection. The experiments show that the performance of this hybrid system is better than that of other techniques regarding to the classification accuracy, the average return and the Sharpe ratio.

Mining User-Generated Contents to Detect Service Failures with Topic Model

Online user-generated contents (UGC) significantly change the way customers behave (e.g., shop, travel), and a pressing need to handle the overwhelmingly plethora amount of various UGC is one of the paramount issues for management. However, a current approach (e.g., sentiment analysis) is often ineffective for leveraging textual information to detect the problems or issues that a certain management suffers from. In this paper, we employ text mining of Latent Dirichlet Allocation (LDA) on a popular online review site dedicated to complaint from users. We find that the employed LDA efficiently detects customer complaints, and a further inspection with the visualization technique is effective to categorize the problems or issues. As such, management can identify the issues at stake and prioritize them accordingly in a timely manner given the limited amount of resources. The findings provide managerial insights into how analytics on social media can help maintain and improve their reputation management. Our interdisciplinary approach also highlights several insights by applying machine learning techniques in marketing research domain. On a broader technical note, this paper illustrates the details of how to implement LDA in R program from a beginning (data collection in R) to an end (LDA analysis in R) since the instruction is still largely undocumented. In this regard, it will help lower the boundary for interdisciplinary researcher to conduct related research.

Automatic Threshold Search for Heat Map Based Feature Selection: A Cancer Dataset Analysis

Public health is one of the most critical issues today; therefore, there is great interest to improve technologies in the area of diseases detection. With machine learning and feature selection, it has been possible to aid the diagnosis of several diseases such as cancer. In this work, we present an extension to the Heat Map Based Feature Selection algorithm, this modification allows automatic threshold parameter selection that helps to improve the generalization performance of high dimensional data such as mass spectrometry. We have performed a comparison analysis using multiple cancer datasets and compare against the well known Recursive Feature Elimination algorithm and our original proposal, the results show improved classification performance that is very competitive against current techniques.

Case-Based Reasoning: A Hybrid Classification Model Improved with an Expert's Knowledge for High-Dimensional Problems

Data mining and classification of objects is the process of data analysis, using various machine learning techniques, which is used today in various fields of research. This paper presents a concept of hybrid classification model improved with the expert knowledge. The hybrid model in its algorithm has integrated several machine learning techniques (Information Gain, K-means, and Case-Based Reasoning) and the expert’s knowledge into one. The knowledge of experts is used to determine the importance of features. The paper presents the model algorithm and the results of the case study in which the emphasis was put on achieving the maximum classification accuracy without reducing the number of features.

An Analysis of Classification of Imbalanced Datasets by Using Synthetic Minority Over-Sampling Technique

Analysing unbalanced datasets is one of the challenges that practitioners in machine learning field face. However, many researches have been carried out to determine the effectiveness of the use of the synthetic minority over-sampling technique (SMOTE) to address this issue. The aim of this study was therefore to compare the effectiveness of the SMOTE over different models on unbalanced datasets. Three classification models (Logistic Regression, Support Vector Machine and Nearest Neighbour) were tested with multiple datasets, then the same datasets were oversampled by using SMOTE and applied again to the three models to compare the differences in the performances. Results of experiments show that the highest number of nearest neighbours gives lower values of error rates. 

Ubiquitous Life People Informatics Engine (U-Life PIE): Wearable Health Promotion System

Since Google launched Google Glass in 2012, numbers of commercial wearable devices were released, such as smart belt, smart band, smart shoes, smart clothes ... etc. However, most of these devices perform as sensors to show the readings of measurements and few of them provide the interactive feedback to the user. Furthermore, these devices are single task devices which are not able to communicate with each other. In this paper a new health promotion system, Ubiquitous Life People Informatics Engine (U-Life PIE), will be presented. This engine consists of People Informatics Engine (PIE) and the interactive user interface. PIE collects all the data from the compatible devices, analyzes this data comprehensively and communicates between devices via various application programming interfaces. All the data and informations are stored on the PIE unit, therefore, the user is able to view the instant and historical data on their mobile devices any time. It also provides the real-time hands-free feedback and instructions through the user interface visually, acoustically and tactilely. These feedback and instructions suggest the user to adjust their posture or habits in order to avoid the physical injuries and prevent illness.

Evaluation of Ensemble Classifiers for Intrusion Detection

One of the major developments in machine learning in the past decade is the ensemble method, which finds highly accurate classifier by combining many moderately accurate component classifiers. In this research work, new ensemble classification methods are proposed with homogeneous ensemble classifier using bagging and heterogeneous ensemble classifier using arcing and their performances are analyzed in terms of accuracy. A Classifier ensemble is designed using Radial Basis Function (RBF) and Support Vector Machine (SVM) as base classifiers. The feasibility and the benefits of the proposed approaches are demonstrated by the means of standard datasets of intrusion detection. The main originality of the proposed approach is based on three main parts: preprocessing phase, classification phase, and combining phase. A wide range of comparative experiments is conducted for standard datasets of intrusion detection. The performance of the proposed homogeneous and heterogeneous ensemble classifiers are compared to the performance of other standard homogeneous and heterogeneous ensemble methods. The standard homogeneous ensemble methods include Error correcting output codes, Dagging and heterogeneous ensemble methods include majority voting, stacking. The proposed ensemble methods provide significant improvement of accuracy compared to individual classifiers and the proposed bagged RBF and SVM performs significantly better than ECOC and Dagging and the proposed hybrid RBF-SVM performs significantly better than voting and stacking. Also heterogeneous models exhibit better results than homogeneous models for standard datasets of intrusion detection. 

Machine Learning Approach for Identifying Dementia from MRI Images

This research paper presents a framework for classifying Magnetic Resonance Imaging (MRI) images for Dementia. Dementia, an age-related cognitive decline is indicated by degeneration of cortical and sub-cortical structures. Characterizing morphological changes helps understand disease development and contributes to early prediction and prevention of the disease. Modelling, that captures the brain’s structural variability and which is valid in disease classification and interpretation is very challenging. Features are extracted using Gabor filter with 0, 30, 60, 90 orientations and Gray Level Co-occurrence Matrix (GLCM). It is proposed to normalize and fuse the features. Independent Component Analysis (ICA) selects features. Support Vector Machine (SVM) classifier with different kernels is evaluated, for efficiency to classify dementia. This study evaluates the presented framework using MRI images from OASIS dataset for identifying dementia. Results showed that the proposed feature fusion classifier achieves higher classification accuracy.

Towards Developing a Self-Explanatory Scheduling System Based on a Hybrid Approach

In the study, we present a conceptual framework for developing a scheduling system that can generate self-explanatory and easy-understanding schedules. To this end, a user interface is conceived to help planners record factors that are considered crucial in scheduling, as well as internal and external sources relating to such factors. A hybrid approach combining machine learning and constraint programming is developed to generate schedules and the corresponding factors, and accordingly display them on the user interface. Effects of the proposed system on scheduling are discussed, and it is expected that scheduling efficiency and system understandability will be improved, compared with previous scheduling systems.

Intelligent Recognition of Diabetes Disease via FCM Based Attribute Weighting

In this paper, an attribute weighting method called fuzzy C-means clustering based attribute weighting (FCMAW) for classification of Diabetes disease dataset has been used. The aims of this study are to reduce the variance within attributes of diabetes dataset and to improve the classification accuracy of classifier algorithm transforming from non-linear separable datasets to linearly separable datasets. Pima Indians Diabetes dataset has two classes including normal subjects (500 instances) and diabetes subjects (268 instances). Fuzzy C-means clustering is an improved version of K-means clustering method and is one of most used clustering methods in data mining and machine learning applications. In this study, as the first stage, fuzzy C-means clustering process has been used for finding the centers of attributes in Pima Indians diabetes dataset and then weighted the dataset according to the ratios of the means of attributes to centers of theirs. Secondly, after weighting process, the classifier algorithms including support vector machine (SVM) and k-NN (k- nearest neighbor) classifiers have been used for classifying weighted Pima Indians diabetes dataset. Experimental results show that the proposed attribute weighting method (FCMAW) has obtained very promising results in the classification of Pima Indians diabetes dataset.