Search results for: supervised machine learning

Authors: Abeer A. Aljohani

Abstract:

COVID-19 virus spread has been one of the extreme pandemics across the globe. It is also referred to as coronavirus, which is a contagious disease that continuously mutates into numerous variants. Currently, the B.1.1.529 variant labeled as omicron is detected in South Africa. The huge spread of COVID-19 disease has affected several lives and has surged exceptional pressure on the healthcare systems worldwide. Also, everyday life and the global economy have been at stake. This research aims to predict COVID-19 disease in its initial stage to reduce the death count. Machine learning (ML) is nowadays used in almost every area. Numerous COVID-19 cases have produced a huge burden on the hospitals as well as health workers. To reduce this burden, this paper predicts COVID-19 disease is based on the symptoms and medical history of the patient. This research presents a unique architecture for COVID-19 detection using ML techniques integrated with feature dimensionality reduction. This paper uses a standard UCI dataset for predicting COVID-19 disease. This dataset comprises symptoms of 5434 patients. This paper also compares several supervised ML techniques to the presented architecture. The architecture has also utilized 10-fold cross validation process for generalization and the principal component analysis (PCA) technique for feature reduction. Standard parameters are used to evaluate the proposed architecture including F1-Score, precision, accuracy, recall, receiver operating characteristic (ROC), and area under curve (AUC). The results depict that decision tree, random forest, and neural networks outperform all other state-of-the-art ML techniques. This achieved result can help effectively in identifying COVID-19 infection cases.

Keywords: supervised machine learning, COVID-19 prediction, healthcare analytics, random forest, neural network

Procedia PDF Downloads 66

Authors: Sadia Nasrin Tisha, Mushfika Sharmin Rahman, Javier Orduz

Abstract:

Machine learning is applied in a variety of fields throughout the world. The healthcare sector has benefited enormously from it. One of the most effective approaches for predicting human heart diseases is to use machine learning applications to classify data and predict the outcome as a classification. However, with the rapid advancement of quantum technology, quantum computing has emerged as a potential game-changer for many applications. Quantum algorithms have the potential to execute substantially faster than their classical equivalents, which can lead to significant improvements in computational performance and efficiency. In this study, we applied quantum machine learning concepts to predict coronary heart diseases from text data. We experimented thrice with three different features; and three feature sets. The data set consisted of 100 data points. We pursue to do a comparative analysis of the two approaches, highlighting the potential benefits of quantum machine learning for predicting heart diseases.

Keywords: quantum machine learning, SVM, QSVM, matrix product state

Procedia PDF Downloads 65

Authors: Azene Zenebe

Abstract:

Deep learning is a subset of machine learning which incorporates techniques for the construction of artificial neural networks and found to be useful for modeling complex problems with large dataset. Deep learning requires a very high power computational and longer time for training. By aggregating computing power, high performance computer (HPC) has emerged as an approach to resolving advanced problems and performing data-driven research activities. Cyber threat intelligence (CIT) is actionable information or insight an organization or individual uses to understand the threats that have, will, or are currently targeting the organization. Results of review of literature will be presented along with results of experimental study that compares the performance of tree-based and function-base machine learning including deep learning algorithms using secondary dataset collected from darknet.

Keywords: deep-learning, cyber security, cyber threat modeling, tree-based machine learning, function-based machine learning, data science

Procedia PDF Downloads 126

Authors: Songul Cinaroglu

Abstract:

Machine learning regression methods are recommended as an alternative to classical regression methods in the existence of variables which are difficult to model. Data for health expenditure is typically non-normal and have a heavily skewed distribution. This study aims to compare machine learning regression methods by hyperparameter tuning to predict health expenditure per capita. A multiple regression model was conducted and performance results of Lasso Regression, Random Forest Regression and Support Vector Machine Regression recorded when different hyperparameters are assigned. Lambda (λ) value for Lasso Regression, number of trees for Random Forest Regression, epsilon (ε) value for Support Vector Regression was determined as hyperparameters. Study results performed by using 'k' fold cross validation changed from 5 to 50, indicate the difference between machine learning regression results in terms of R², RMSE and MAE values that are statistically significant (p < 0.001). Study results reveal that Random Forest Regression (R² ˃ 0.7500, RMSE ≤ 0.6000 ve MAE ≤ 0.4000) outperforms other machine learning regression methods. It is highly advisable to use machine learning regression methods for modelling health expenditures.

Keywords: machine learning, lasso regression, random forest regression, support vector regression, hyperparameter tuning, health expenditure

Procedia PDF Downloads 195

Authors: Suraj Gururaj, Sumantha Udupa U.

Abstract:

Browsing the Web has emerged as the de facto activity performed on the Internet. Although browsing gets tracked, the manageability aspect of Web browsing history is very poor. In this paper, we have a workable solution implemented by using machine learning and natural language processing techniques for efficient manageability of user’s browsing history. The significance of adding such a capability to a Web browser is that it ensures efficient and quick information retrieval from browsing history, which currently is very challenging. Our solution guarantees that any important websites visited in the past can be easily accessible because of the intelligent and automatic classification. In a nutshell, our solution-based paper provides an implementation as a browser extension by intelligently classifying the browsing history into most relevant category automatically without any user’s intervention. This guarantees no information is lost and increases productivity by saving time spent revisiting websites that were of much importance.

Keywords: adhoc retrieval, Chrome extension, supervised learning, tile, Web personalization

Procedia PDF Downloads 347

Authors: Sam Khozama, Ali M. Mayya

Abstract:

Breast cancer is one of the most common types in women. Early prediction of breast cancer helps physicians detect cancer in its early stages. Big cancer data needs a very powerful tool to analyze and extract predictions. Machine learning and deep learning are two of the most efficient tools for predicting cancer based on textual data. In this study, we developed a fusion model of two machine learning and deep learning models. To obtain the final prediction, Long-Short Term Memory (LSTM) and ensemble learning with hyper parameters optimization are used, and score-level fusion is used. Experiments are done on the Breast Cancer Surveillance Consortium (BCSC) dataset after balancing and grouping the class categories. Five different training scenarios are used, and the tests show that the designed fusion model improved the performance by 3.3% compared to the individual models.

Keywords: machine learning, deep learning, cancer prediction, breast cancer, LSTM, fusion

Procedia PDF Downloads 135

Authors: David Ming Liu, Benjamin Hirsch, Bashir Aden

Abstract:

Syndromic surveillance is to detect or predict disease outbreaks through the analysis of medical sources of data. Using social media data like tweets to do syndromic surveillance becomes more and more popular with the aid of open platform to collect data and the advantage of microblogging text and mobile geographic location features. In this paper, a Syndromic Surveillance Framework is presented with machine learning kernel using tweets data analytics. Influenza and the three cities Abu Dhabi, Al Ain and Dubai of United Arabic Emirates are used as the test disease and trial areas. Hospital cases data provided by the Health Authority of Abu Dhabi (HAAD) are used for the correlation purpose. In our model, Latent Dirichlet allocation (LDA) engine is adapted to do supervised learning classification and N-Fold cross validation confusion matrix are given as the simulation results with overall system recall 85.595% performance achieved.

Keywords: Syndromic surveillance, Tweets, Machine Learning, data mining, Latent Dirichlet allocation (LDA), Influenza

Procedia PDF Downloads 89

Authors: Vincent Liu

Abstract:

Machine learning algorithms (ML) can achieve high accuracy in predicting outcomes compared to classical models. Metaheuristic, nature-inspired algorithms can enhance traditional ML algorithms by optimizing them such as by performing feature selection. We compare ten ML algorithms to predict 30-day hospital readmission rates for diabetes patients in the US using a dataset from UCI Machine Learning Repository with feature selection performed by Greywolf nature-inspired algorithm. The baseline accuracy for the initial random forest model was 65%. After performing feature engineering, SMOTE for class balancing, and Greywolf optimization, the machine learning algorithms showed better metrics, including F1 scores, accuracy, and confusion matrix with improvements ranging in 10%-30%, and a best model of XGBoost with an accuracy of 95%. Applying machine learning this way can improve patient outcomes as unnecessary rehospitalizations can be prevented by focusing on patients that are at a higher risk of readmission.

Keywords: diabetes, machine learning, 30-day readmission, metaheuristic

Procedia PDF Downloads 28

Authors: Rodrigo Aguiar, Adelino Ferreira

Abstract:

Road traffic accidents are the leading cause of unnatural death and injuries worldwide, representing a significant problem of road safety. In this context, the use of artificial intelligence with advanced machine learning techniques has gained prominence as a promising approach to predict traffic accidents. This article investigates the application of machine learning algorithms to develop traffic accident frequency prediction models. Models are evaluated based on performance metrics, making it possible to do a comparative analysis with traditional prediction approaches. The results suggest that machine learning can provide a powerful tool for accident prediction, which will contribute to making more informed decisions regarding road safety.

Keywords: machine learning, artificial intelligence, frequency of accidents, road safety

Procedia PDF Downloads 58

Authors: Seongjun Kim, Sanghoon Shim, Jinwooung Kim, Jaehwan Jung, Sung-Ah Kim

Abstract:

As the development of information and communication technology, the convergence of machine learning of the ICT area and design is attempted. In this way, it is possible to grasp the correlation between various design elements, which was difficult to grasp, and to reflect this in the design result. In architecture, there is an attempt to predict the performance, which is difficult to grasp in the past, by finding the correlation among multiple factors mainly through machine learning. In architectural design area, some attempts to predict the performance affected by various factors have been tried. With machine learning, it is possible to quickly predict performance. The aim of this study is to propose a model that predicts performance according to the block arrangement of apartment housing through machine learning and the design alternative which satisfies the performance such as the daylight hours in the most similar form to the alternative proposed by the designer. Through this study, a designer can proceed with the design considering various design alternatives and accurate performances quickly from the early design stage.

Keywords: apartment housing, machine learning, multi-objective optimization, performance prediction

Procedia PDF Downloads 451

Authors: Sajjad Ahmed, Knut Hinkelmann, Flavio Corradini

Abstract:

Fake news detection research is still in the early stage as this is a relatively new phenomenon in the interest raised by society. Machine learning helps to solve complex problems and to build AI systems nowadays and especially in those cases where we have tacit knowledge or the knowledge that is not known. We used machine learning algorithms and for identification of fake news; we applied three classifiers; Passive Aggressive, Naïve Bayes, and Support Vector Machine. Simple classification is not completely correct in fake news detection because classification methods are not specialized for fake news. With the integration of machine learning and text-based processing, we can detect fake news and build classifiers that can classify the news data. Text classification mainly focuses on extracting various features of text and after that incorporating those features into classification. The big challenge in this area is the lack of an efficient way to differentiate between fake and non-fake due to the unavailability of corpora. We applied three different machine learning classifiers on two publicly available datasets. Experimental analysis based on the existing dataset indicates a very encouraging and improved performance.

Keywords: fake news detection, natural language processing, machine learning, classification techniques.

Procedia PDF Downloads 135

Authors: Joseph Salim

Abstract:

This paper proposes the development of a supervised machine learning system to predict the consequences of untreated bruxism (teeth grinding) on gingival (gum) recession and jaw pain (most often bilateral jaw pain with possible headaches and limited ability to open the mouth). As a general dentist in a multi-specialty practice, the author has encountered many patients suffering from these issues due to uncontrolled bruxism (teeth grinding) at night. The most effective treatment for managing this problem involves wearing a nightguard during sleep and receiving therapeutic Botox injections to relax the muscles (the masseter muscle) responsible for grinding. However, some patients choose to postpone these treatments, leading to potentially irreversible and costlier consequences in the future. The proposed machine learning model aims to track patients who forgo the recommended treatments and assess the percentage of individuals who will experience worsening jaw pain, gingival (gum) recession, or both within a 3-to-5-year timeframe. By accurately predicting these outcomes, the model seeks to motivate patients to address the root cause proactively, ultimately saving time and pain while improving quality of life and avoiding much costlier treatments such as full-mouth rehabilitation to help recover the loss of vertical dimension of occlusion due to shortened clinical crowns because of bruxism, gingival grafts, etc.

Keywords: artificial intelligence, machine learning, predictive insights, bruxism, teeth grinding, therapeutic botox, nightguard, gingival recession, gum recession, jaw pain

Procedia PDF Downloads 60

Authors: Radhika Ranjan Roy

Abstract:

Deployment of machine learning (ML)/deep learning (DL) algorithms for cyberattack detection in operational communications networks (wireless and/or wire-line) is being delayed because of low-performance parameters (e.g., recall, precision, and f₁-score). If datasets become imbalanced, which is the usual case for communications networks, the performance tends to become worse. Complexities in handling reducing dimensions of the feature sets for increasing performance are also a huge problem. Mahalanobis algorithms have been widely applied in scientific research because Mahalanobis distance metric learning is a successful framework. In this paper, we have investigated the Mahalanobis binary classifier algorithm for increasing cyberattack detection performance over communications networks as a proof of concept. We have also found that high-dimensional information in intermediate features that are not utilized as much for classification tasks in ML/DL algorithms are the main contributor to the state-of-the-art of improved performance of the Mahalanobis method, even for imbalanced and sparse datasets. With no feature reduction, MD offers uniform results for precision, recall, and f₁-score for unbalanced and sparse NSL-KDD datasets.

Keywords: Mahalanobis distance, machine learning, deep learning, NS-KDD, local intrinsic dimensionality, chi-square, positive semi-definite, area under the curve

Procedia PDF Downloads 52

Authors: Daniel M. Wolff, Patricia Gray, Rafael de la Parra Venegas

Abstract:

An unsupervised machine listening system is constructed and applied to a dataset of 17,195 30-second marine hydrophone recordings. The system is then heuristically supplemented with anecdotal listening, contextual recording information, and supervised learning techniques to reduce the number of false positives. Features for classification are assembled by extracting the following data from each of the audio files: the spectral centroid, root-mean-squared values for each frequency band of a 10-octave filter bank, and mel-frequency cepstral coefficients in 5-second frames. In this way both time- and frequency-domain information are contained in the features to be passed to a clustering algorithm. Classification is performed using the k-means algorithm and then a k-nearest neighbors search. Different values of k are experimented with, in addition to different combinations of the available feature sets. Hypothesized class labels are 'primarily anthrophony' and 'primarily biophony', where the best class result conforming to the former label has 104 members after heuristic pruning. This demonstrates how a large audio dataset has been made more tractable with machine learning techniques, forming the foundation of a framework designed to acoustically monitor and gauge biological and anthropogenic activity in a marine environment.

Keywords: anthrophony, hydrophone, k-means, machine learning

Procedia PDF Downloads 136

Authors: Carlos Roncero-Parra, Alfonso Parreño-Torres, Jorge Mateo Sotos, Alejandro L. Borja

Abstract:

In this research, different methods based on machine/deep learning algorithms are presented for the classification and diagnosis of patients with mental disorders such as alzheimer. For this purpose, the signals obtained from 32 unipolar electrodes identified by non-invasive EEG were examined, and their basic properties were obtained. More specifically, different well-known machine learning based classifiers have been used, i.e., support vector machine (SVM), Bayesian linear discriminant analysis (BLDA), decision tree (DT), Gaussian Naïve Bayes (GNB), K-nearest neighbor (KNN) and Convolutional Neural Network (CNN). A total of 668 patients from five different hospitals have been studied in the period from 2011 to 2021. The best accuracy is obtained was around 93 % in both ADM and ADA classifications. It can be concluded that such a classification will enable the training of algorithms that can be used to identify and classify different mental disorders with high accuracy.

Keywords: alzheimer, machine learning, deep learning, EEG

Procedia PDF Downloads 88

Authors: Anli Du Preez, James Bekker

Abstract:

Data is currently one of the most critical and influential emerging technologies. However, the true potential of data is yet to be exploited since, currently, about 1% of generated data are ever actually analyzed for value creation. There is a data gap where data is not explored due to the lack of data analytics infrastructure and the required data analytics skills. This study developed a decision support framework for data analytics by following Jabareen’s framework development methodology. The study focused on machine learning algorithms, which is a subset of data analytics. The developed framework is designed to assist data analysts with little experience, in choosing the appropriate machine learning algorithm given the purpose of their application.

Keywords: Data analytics, Industrial engineering, Machine learning, Value creation

Procedia PDF Downloads 144

Authors: Peter Adekunle, Clinton Aigbavboa, Matthew Ikuabe, Opeoluwa Akinradewo

Abstract:

The future of machine learning (ML) in building may seem like a distant idea that will take decades to materialize, but it is actually far closer than previously believed. In reality, the built environment has been progressively increasing interest in machine learning. Although it could appear to be a very technical, impersonal approach, it can really make things more personable. Instead of eliminating humans out of the equation, machine learning allows people do their real work more efficiently. It is therefore vital to evaluate the factors influencing the implementation and challenges of implementing machine learning techniques in the South African built environment. The study's design was one of a survey. In South Africa, construction workers and professionals were given a total of one hundred fifty (150) questionnaires, of which one hundred and twenty-four (124) were returned and deemed eligible for study. Utilizing percentage, mean item scores, standard deviation, and Kruskal-Wallis, the collected data was analyzed. The results demonstrate that the top factors influencing the adoption of machine learning are knowledge level and a lack of understanding of its potential benefits. While lack of collaboration among stakeholders and lack of tools and services are the key hurdles to the deployment of machine learning within the South African built environment. The study came to the conclusion that ML adoption should be promoted in order to increase safety, productivity, and service quality within the built environment.

Keywords: machine learning, implementation, built environment, construction stakeholders

Procedia PDF Downloads 105

Authors: Nabil Shawkat, Jamil Saquer

Abstract:

Social media has accelerated our ability to engage with others and eliminated many communication barriers. On the other hand, the widespread use of social media resulted in an increase in online hate speech. This has drastic impacts on vulnerable individuals and societies. Therefore, it is critical to detect hate speech to prevent innocent users and vulnerable communities from becoming victims of hate speech. We investigate the performance of different deep learning and machine learning algorithms on three different datasets. Our results show that the BERT model gives the best performance among all the models by achieving an F1-score of 90.6% on one of the datasets and F1-scores of 89.7% and 88.2% on the other two datasets.

Keywords: hate speech, machine learning, deep learning, abusive words, social media, text classification

Procedia PDF Downloads 108

Authors: Dim En Nyaung, Thin Lai Lai Thein

Abstract:

Due to the rapid increase of Internet, web opinion sources dynamically emerge which is useful for both potential customers and product manufacturers for prediction and decision purposes. These are the user generated contents written in natural languages and are unstructured-free-texts scheme. Therefore, opinion mining techniques become popular to automatically process customer reviews for extracting product features and user opinions expressed over them. Since customer reviews may contain both opinionated and factual sentences, a supervised machine learning technique applies for subjectivity classification to improve the mining performance. In this paper, we dedicate our work is the task of opinion summarization. Therefore, product feature and opinion extraction is critical to opinion summarization, because its effectiveness significantly affects the identification of semantic relationships. The polarity and numeric score of all the features are determined by Senti-WordNet Lexicon. The problem of opinion summarization refers how to relate the opinion words with respect to a certain feature. Probabilistic based model of supervised learning will improve the result that is more flexible and effective.

Keywords: opinion mining, opinion summarization, sentiment analysis, text mining

Procedia PDF Downloads 309

Authors: Vincenzo Capalbo, Gianpaolo Ghiani, Emanuele Manni

Abstract:

Global e-commerce sales have reached unprecedented levels in the past few years. As this trend is only predicted to go up as we continue into the ’20s, new challenges will be faced by companies when planning and controlling e-commerce logistics. In this paper, we survey the related literature on Optimization and Machine Learning as well as on combined methodologies. We also identify the distinctive features of next-generation planning algorithms - namely scalability, model-and-run features and learning capabilities - that will be fundamental to cope with the scale and complexity of logistics in the next decade.

Keywords: e-commerce, hardware acceleration, logistics, machine learning, mixed integer programming, optimization

Procedia PDF Downloads 210

Authors: Theoktisti Makridou, Konstantinos Tsaprailis, George Arvanitakis, Charalampos Kontoes

Abstract:

In the literature, the data-driven approaches for mosquito abundance prediction relaying on supervised machine learning models that get trained with historical in-situ measurements. The counterpart of this approach is once the model gets trained on pointlevel (specific x,y coordinates) measurements, the predictions of the model refer again to point-level. These point-level predictions reduce the applicability of those solutions once a lot of early warning and mitigation actions applications need predictions for an area level, such as a municipality, village, etc... In this study, we apply a data-driven predictive model, which relies on public-open satellite Earth Observation and geospatial data and gets trained with historical point-level in-Situ measurements of mosquito abundance. Then we propose a methodology to extract information from a point-level predictive model to a broader area-level prediction. Our methodology relies on the randomly spatial sampling of the area of interest (similar to the Poisson hardcore process), obtaining the EO and geomorphological information for each sample, doing the point-wise prediction for each sample, and aggregating the predictions to represent the average mosquito abundance of the area. We quantify the performance of the transformation from the pointlevel to the area-level predictions, and we analyze it in order to understand which parameters have a positive or negative impact on it. The goal of this study is to propose a methodology that predicts the mosquito abundance of a given area by relying on point-level prediction and to provide qualitative insights regarding the expected performance of the area-level prediction. We applied our methodology to historical data (of Culex pipiens) of two areas of interest (Veneto region of Italy and Central Macedonia of Greece). In both cases, the results were consistent. The mean mosquito abundance of a given area can be estimated with similar accuracy to the point-level predictor, sometimes even better. The density of the samples that we use to represent one area has a positive effect on the performance in contrast to the actual number of sampling points which is not informative at all regarding the performance without the size of the area. Additionally, we saw that the distance between the sampling points and the real in-situ measurements that were used for training did not strongly affect the performance.

Keywords: mosquito abundance, supervised machine learning, culex pipiens, spatial sampling, west nile virus, earth observation data

Procedia PDF Downloads 112

Authors: Pawel Twardowski, Maciej Tabaszewski, Jakub Czyżycki

Abstract:

The main goal of the work was to use machine learning to predict cutting-edge wear. The research was carried out while milling hardened steel with sintered carbide cutters at various cutting speeds. During the tests, cutting-edge wear was measured, and vibration acceleration signals were also measured. Appropriate measures were determined from the vibration signals and served as input data in the machine-learning process. Two approaches were used in this work. The first one involved a two-state classification of the cutting edge - suitable and unfit for further work. In the second approach, prediction of the cutting-edge state based on vibration signals was used. The obtained research results show that the appropriate use of machine learning algorithms gives excellent results related to monitoring cutting edge during the process.

Keywords: milling of hardened steel, tool wear, vibrations, machine learning

Procedia PDF Downloads 24

Authors: Yizhen Zhao, Adam S. Z. Belloum, Goncalo Maia Da Costa, Zhiming Zhao

Abstract:

Machine learning has evolved from an area of academic research to a real-word applied field. This change comes with challenges, gaps and differences exist between common practices in academic environments and the ones in production environments. Following continuous integration, development and delivery practices in software engineering, similar trends have happened in machine learning (ML) systems, called MLOps. In this paper we propose a framework that helps to streamline and introduce best practices that facilitate the ML lifecycle in an industrial setting. This framework can be used as a template that can be customized to implement various machine learning experiment. The proposed framework is modular and can be recomposed to be adapted to various use cases (e.g. data versioning, remote training on cloud). The framework inherits practices from DevOps and introduces other practices that are unique to the machine learning system (e.g.data versioning). Our MLOps practices automate the entire machine learning lifecycle, bridge the gap between development and operation.

Keywords: cloud computing, continuous development, data versioning, DevOps, industrial setting, MLOps

Procedia PDF Downloads 237

Authors: Talal Alshammari, Nasser Alshammari, Mohamed Sedky, Chris Howard

Abstract:

With the widespread adoption of the Internet-connected devices, and with the prevalence of the Internet of Things (IoT) applications, there is an increased interest in machine learning techniques that can provide useful and interesting services in the smart home domain. The areas that machine learning techniques can help advance are varied and ever-evolving. Classifying smart home inhabitants’ Activities of Daily Living (ADLs), is one prominent example. The ability of machine learning technique to find meaningful spatio-temporal relations of high-dimensional data is an important requirement as well. This paper presents a comparative evaluation of state-of-the-art machine learning techniques to classify ADLs in the smart home domain. Forty-two synthetic datasets and two real-world datasets with multiple inhabitants are used to evaluate and compare the performance of the identified machine learning techniques. Our results show significant performance differences between the evaluated techniques. Such as AdaBoost, Cortical Learning Algorithm (CLA), Decision Trees, Hidden Markov Model (HMM), Multi-layer Perceptron (MLP), Structured Perceptron and Support Vector Machines (SVM). Overall, neural network based techniques have shown superiority over the other tested techniques.

Keywords: activities of daily living, classification, internet of things, machine learning, prediction, smart home

Procedia PDF Downloads 320

Authors: B. R. Campomanes-Alvarez, P. Quiros, B. Fernandez

Abstract:

Automatic Speech Recognition (ASR) is a machine-based process of decoding and transcribing oral speech. A typical ASR system receives acoustic input from a speaker or an audio file, analyzes it using algorithms, and produces an output in the form of a text. Some speech recognition systems use Hidden Markov Models (HMMs) to deal with the temporal variability of speech and Gaussian Mixture Models (GMMs) to determine how well each state of each HMM fits a short window of frames of coefficients that represents the acoustic input. Another way to evaluate the fit is to use a feed-forward neural network that takes several frames of coefficients as input and produces posterior probabilities over HMM states as output. Deep neural networks (DNNs) that have many hidden layers and are trained using new methods have been shown to outperform GMMs on a variety of speech recognition systems. Acoustic models for state-of-the-art ASR systems are usually training on massive amounts of data. However, audio files with their corresponding transcriptions can be difficult to obtain, especially in the Spanish language. Hence, in the case of these low-resource scenarios, building an ASR model is considered as a complex task due to the lack of labeled data, resulting in an under-trained system. Semi-supervised learning approaches arise as necessary tasks given the high cost of transcribing audio data. The main goal of this proposal is to develop a procedure based on acoustic semi-supervised learning for Spanish ASR systems by using DNNs. This semi-supervised learning approach consists of: (a) Training a seed ASR model with a DNN using a set of audios and their respective transcriptions. A DNN with a one-hidden-layer network was initialized; increasing the number of hidden layers in training, to a five. A refinement, which consisted of the weight matrix plus bias term and a Stochastic Gradient Descent (SGD) training were also performed. The objective function was the cross-entropy criterion. (b) Decoding/testing a set of unlabeled data with the obtained seed model. (c) Selecting a suitable subset of the validated data to retrain the seed model, thereby improving its performance on the target test set. To choose the most precise transcriptions, three confidence scores or metrics, regarding the lattice concept (based on the graph cost, the acoustic cost and a combination of both), was performed as selection technique. The performance of the ASR system will be calculated by means of the Word Error Rate (WER). The test dataset was renewed in order to extract the new transcriptions added to the training dataset. Some experiments were carried out in order to select the best ASR results. A comparison between a GMM-based model without retraining and the DNN proposed system was also made under the same conditions. Results showed that the semi-supervised ASR-model based on DNNs outperformed the GMM-model, in terms of WER, in all tested cases. The best result obtained an improvement of 6% relative WER. Hence, these promising results suggest that the proposed technique could be suitable for building ASR models in low-resource environments.

Keywords: automatic speech recognition, deep neural networks, machine learning, semi-supervised learning

Procedia PDF Downloads 320

Authors: Vishnu Pratap Singh Kirar, Madhuri Saxena

Abstract:

Advancement in sequence data generation technologies is churning out voluminous omics data and posing a massive challenge to annotate the biological functional features. With so much data available, the use of machine learning methods and tools to make novel inferences has become obvious. Machine learning methods have been successfully applied to a lot of disciplines, including computational biology and bioinformatics. Researchers in computational biology are interested to develop novel machine learning frameworks to classify the huge amounts of biological data. In this proposal, it plan to employ novel machine learning approaches to aid the understanding of how apparently innocuous mutations (in intergenic DNA and at synonymous sites) cause diseases. We are also interested in discovering novel functional sites in the genome and mutations in which can affect a phenotype of interest.

Keywords: genome wide association studies (GWAS), next generation sequencing (NGS), deep learning, omics

Procedia PDF Downloads 67

Authors: Shuen-Tai Wang, Fang-An Kuo, Chau-Yi Chou, Yu-Bin Fang

Abstract:

2016 has become the year of the Artificial Intelligence explosion. AI technologies are getting more and more matured that most world well-known tech giants are making large investment to increase the capabilities in AI. Machine learning is the science of getting computers to act without being explicitly programmed, and deep learning is a subset of machine learning that uses deep neural network to train a machine to learn  features directly from data. Deep learning realizes many machine learning applications which expand the field of AI. At the present time, deep learning frameworks have been widely deployed on servers for deep learning applications in both academia and industry. In training deep neural networks, there are many standard processes or algorithms, but the performance of different frameworks might be different. In this paper we evaluate the running performance of two state-of-the-art distributed deep learning frameworks that are running training calculation in parallel over multi GPU and multi nodes in our cloud environment. We evaluate the training performance of the frameworks with ResNet-50 convolutional neural network, and we analyze what factors that result in the performance among both distributed frameworks as well. Through the experimental analysis, we identify the overheads which could be further optimized. The main contribution is that the evaluation results provide further optimization directions in both performance tuning and algorithmic design.

Keywords: artificial intelligence, machine learning, deep learning, convolutional neural networks

Procedia PDF Downloads 178

Authors: Hailyie Tekleselase

Abstract:

Deep learning is increasingly used as a building block of security systems. However, neural networks are hard to interpret and typically solid to the practitioner. This paper presents a detail survey of computing methods in cyber security, and analyzes the prospects of enhancing the cyber security capabilities by suggests that of accelerating the intelligence of the security systems. There are many AI-based applications used in industrial scenarios such as Internet of Things (IoT), smart grids, and edge computing. Machine learning technologies require a training process which introduces the protection problems in the training data and algorithms. We present machine learning techniques currently applied to the detection of intrusion, malware, and spam. Our conclusions are based on an extensive review of the literature as well as on experiments performed on real enterprise systems and network traffic. We conclude that problems can be solved successfully only when methods of artificial intelligence are being used besides human experts or operators.

Keywords: artificial intelligence, machine learning, deep learning, cyber security, big data

Procedia PDF Downloads 102

Authors: Jun Wang, Ge Zhang

Abstract:

Machine learning has been used in many areas today. It thrives at reviewing large volumes of data and identifying patterns and trends that might not be apparent to a human. Given the huge potential benefit and the amount of data available in the financial market, it is not surprising to see machine learning applied to various financial products. While future prices of financial securities are extremely difficult to forecast, we study them from a different angle. Instead of trying to forecast future prices, we apply machine learning algorithms to predict the direction of future price movement, in particular, whether a sector Exchange Traded Fund (ETF) would outperform or underperform the market in the next week or in the next month. We apply several machine learning algorithms for this prediction. The algorithms are Linear Discriminant Analysis (LDA), k-Nearest Neighbors (KNN), Decision Tree (DT), Gaussian Naive Bayes (GNB), and Neural Networks (NN). We show that these machine learning algorithms, most notably GNB and NN, have some predictive power in forecasting out-performance and under-performance out of sample. We also try to explore whether it is possible to utilize the predictions from these algorithms to outperform the buy-and-hold strategy of the S&P 500 index. The trading strategy to explore out-performance predictions does not perform very well, but the trading strategy to explore under-performance predictions can earn higher returns than simply holding the S&P 500 index out of sample.

Keywords: machine learning, ETF prediction, dynamic trading, asset allocation

Procedia PDF Downloads 63

Authors: Newton Muhury, Armando A. Apan, Tek N. Marasani, Gebiaw T. Ayele

Abstract:

The changing climate has degraded freshwater availability in Australia that influencing vegetation growth to a great extent. This study assessed the vegetation responses to groundwater using Terra’s moderate resolution imaging spectroradiometer (MODIS), Normalised Difference Vegetation Index (NDVI), and soil water content (SWC). A hydrological model, SWAT, has been set up in a southeast Australian river catchment for groundwater analysis. The model was calibrated and validated against monthly streamflow from 2001 to 2006 and 2007 to 2010, respectively. The SWAT simulated soil water content for 43 sub-basins and monthly MODIS NDVI data for three different types of vegetation (forest, shrub, and grass) were applied in the machine learning tool, Waikato Environment for Knowledge Analysis (WEKA), using two supervised machine learning algorithms, i.e., support vector machine (SVM) and random forest (RF). The assessment shows that different types of vegetation response and soil water content vary in the dry and wet seasons. The WEKA model generated high positive relationships (r = 0.76, 0.73, and 0.81) between NDVI values of all vegetation in the sub-basins against soil water content (SWC), the groundwater flow (GW), and the combination of these two variables, respectively, during the dry season. However, these responses were reduced by 36.8% (r = 0.48) and 13.6% (r = 0.63) against GW and SWC, respectively, in the wet season. Although the rainfall pattern is highly variable in the study area, the summer rainfall is very effective for the growth of the grass vegetation type. This study has enriched our knowledge of vegetation responses to groundwater in each season, which will facilitate better floodplain vegetation management.

Keywords: ArcSWAT, machine learning, floodplain vegetation, MODIS NDVI, groundwater

Procedia PDF Downloads 70