Hidden Markov Model Model Research Articles

KiNext: a portable and scalable workflow for the identification and classification of protein kinases

BackgroundProtein kinases are a diverse superfamily of proteins common to organisms across the tree of life that are typically involved in signal transduction, allowing organisms to sense and respond to biotic or abiotic environmental factors. They have important roles in organismal physiology, including development, reproduction, acclimation to environmental stress, while their dysregulation can lead to disease, including several forms of cancer. Identifying the complement of protein kinases (the kinome) of any organism is useful for understanding its physiological capabilities, limitations and adaptations to environmental stress. The increasing availability of genomes makes it now possible to examine and compare the kinomes across a broad diversity of organisms. Here we present a pipeline respecting the FAIR principles (findable, accessible, interoperable and reusable) that facilitates the search and identification of protein kinases from a predicted proteome, and classifies them according to group of serine/threonine/tyrosine protein kinases present in eukaryotes.ResultsKiNext is a Nextflow pipeline that regroups a number of existing bioinformatic tools to search for and classify the protein kinases of an organism in a reproducible manner, starting from a set of amino acid sequences. Conventional eukaryotic protein kinases (ePKs) and atypical protein kinases (aPKs) are identified by using Hidden Markov Models (HMMs) generated from the catalytic domains of kinases. Furthermore, KiNext categorizes ePKs into the eight kinase groups by employing dedicated Hidden Markov Models (HMMs) tailored for each group. The performance of the KiNext pipeline was validated against previously identified kinomes obtained with other tools that were already published for two marine species, the Pacific oyster Crassostrea gigas and the unicellular green alga Ostreoccocus tauri. KiNext outperformed previous results by finding previously unidentified kinases and by attributing a large proportion of previously unclassified kinases to a group in both species. These results demonstrate improvements in kinase identification and classification, all while providing traceability and reproducibility of results in a FAIR pipeline. The default HMM models provided with KiNext are most suitable for eukaryotes, but the pipeline can be easily modified to include HMM models for other taxa of interest.ConclusionThe KiNext pipeline enables efficient and reproducible identification of kinomes based on predicted amino acid sequences (i.e. proteomes). KiNext was designed to be easy to use, automated, portable and scalable.

Energy demand pattern analysis in South Korea using hidden Markov model‐based classification

AbstractUnderstanding energy demand patterns in the residential sector is crucial for improving energy efficiency through demand‐side management. Load curve classification is a useful method for analyzing energy demand patterns. In this paper, we employ a hidden Markov model (HMM)‐based classification to residential load curves in South Korea. We also investigate how the number of hidden states affects classification performance by allowing HMM to train with a different number of hidden states for each class. We compare our HMM‐based method with several state‐of‐the‐art models and find that it outperforms other competing models in multiple datasets. Additionally, we use the fitted HMM model to make inferences about the load curves, gaining deeper insights into energy demand patterns.

Handwritten Online Signature Verification and Forgery Detection Using Hybrid Wavelet Transform-2 with HMM Classifier

Online signature verification stands out as a distinctive biometric feature, offering both static and dynamic attributes within 2D signature images. A Hybrid Wavelet Transform-2 (HWT-2) with a size of 256 is constructed by employing the Kronecker product of two orthogonal transforms: DCT, DHT, Haar, Hadamard, and Kekre, each with sizes of 4 and 64. The HWT enables the analysis of signals at both global and local levels, akin to wavelet transforms. HWT-2 is applied to 256 samples of online handwritten signatures, and the first 128 samples of the output are utilized as feature vectors for the verification and forgery detection of online handwritten signatures. These feature vectors are inputted into Left-Right and Ergodic Hidden Markov Model (HMM) classifiers for analysis. The HMMs are trained using 10 randomly selected genuine signature samples, and subsequently tested on the remaining 10 genuine signatures and 20 forged signatures from 40 users of the SVC 2004 signature database. This process is repeated 20 times, and the average values are computed. Among all possible combinations of HWT-2 using DCT, DHT, Haar, Hadamard, and Kekre transforms for the Left-Right HMM model, the combination of DCT 4 and DHT 64 demonstrates the best performance, with False Rejection Rate (FRR) and False Acceptance Rate (FAR) values of 3.96% and 1.48%, respectively, for state 5. Similarly, for the Ergodic HMM model, the combination of DCT 4 and DHT 64 exhibits the best performance, with FRR and FAR values of 1.10% and 2.88%, respectively, for state 5. These results indicate that combinations of HWT-2 outperform individual orthogonal transforms, and further, that HWT-2 combinations within the Ergodic HMM model offer superior performance compared to the Left-Right HMM model.

Handwritten online signature verification and Forgery Detection for Online Handwritten Signature using Hybrid Wavelet Transform-1 with HMM classifier

Online signature verification employs a distinctive biometric trait by utilizing both static and dynamic features extracted from 2D signature images. A hybrid wavelet transform, denoted as HWT-1 with a size of 256, is formed through the Kronecker product of two orthogonal transforms, such as DCT, DHT, Haar, Hadamard, and Kekre, each with sizes 4 and 64. This HWT facilitates signal analysis at both global and local levels, akin to traditional wavelet transforms. Specifically, HWT-1 processes the 256 samples of online handwritten signatures, yielding 128 samples that constitute the feature vectors for signature verification and forgery detection. These feature vectors are then inputted into the Left-Right and Ergodic Hidden Markov Model (HMM) classifiers for further analysis. The HMMs are trained using 10 randomly selected genuine signature samples and subsequently tested with the remaining 10 genuine signatures and 20 forged signatures from the SVC 2004 signature database, repeating this process 20 times to compute average values. Among all possible combinations of HWT-1 utilizing DCT, DHT, Haar, Hadamard, and Kekre transforms for the Left-Right HMM model, the combination of DCT 4 and Haar 64 demonstrates the best performance with a False Rejection Rate (FRR) and False Acceptance Rate (FAR) of 1.05% and 0.99%, respectively, at state 3. Similarly, considering all feasible combinations of HWT-1 for the Ergodic HMM model, the combination of DCT 4 and Kekre 64 yields the optimal performance with an FRR and FAR of 1.24% and 1.33%, respectively, at state 3.

HMM-Based Method for Aircraft Environmental Control System Turbofan Rolling Bearing Fault Diagnosis

In response to the high-noise, nonlinear, and nonstationary characteristics of vibration signals from aircraft environmental control system (ECS) turbofan rolling bearings, this paper proposes a diagnostic method for the degree of ECS turbofan bearing faults based on the Hidden Markov Model (HMM). Experimental results demonstrate that HMM can accurately diagnose and predict faults in ECS turbofan rolling bearings. The HMM method enhances diagnostic accuracy, and its effectiveness and feasibility in fault diagnosis based on different rolling bearing fault instances are elaborated. By employing the HMM model to establish precise models from decomposed dynamic data, it successfully identifies faults such as the fracture of the bearing cage under biased load conditions, although its performance in recognizing overheating faults is suboptimal.

Exploration of Intelligent Transformation Path of Traditional Cultural and Creative Products Based on Internet of Things Technology

Abstract More and more modern technology is applied to traditional cultural and creative industries, which enhances the technicality and novelty of traditional cultural products, and the fusion of culture and modern technology profoundly changes people’s way of life and learning. This paper explores the Internet of Things (IoT) technology and the intelligent transformation process of cultural and creative products and constructs cultural and creative products that integrate intelligent voice series. Specifically, a speech recognition model is integrated into the design process of cultural and creative products, and the designed model is applied to the design of intelligent cultural and creative products by analyzing the Hidden Markov Model (HMM), proposing a solution to the shortcomings of the HMM model, and choosing to perform noise reduction of speech signals through the fixed beam algorithm. Based on the speech recognition model designed in this paper, an intelligent cultural and creative product with a speech recognition function is designed, taking the ‘owl wine container’ from the Shanxi Museum as an example. The regression equation of the audience’s satisfaction with the product is Y (satisfaction) = -0.000263 + 0.208X1 (practicality) + 0.265X2 (innovation) + 0.253X3 (culture) + 0.271X4 (interactivity) - 0.296X5 (fun). This paper’s intelligent cultural and creative products have a greater impact on satisfaction due to their innovativeness and interactive function.

A map matching algorithm based on modified hidden Markov model considering time series dependency over larger time span

With the advancement of geopositioning systems and mobile devices, much research with geopositioning data are currently ongoing. Along with the research applications, map matching is a technology that infers the actual position of error-prone trajectory data. It is a core preprocessing technique for trajectory data. Among various map matching algorithms, map matching using Hidden Markov Model (HMM) has gained high attention. However, the HMM model simplifies the dependency of time series data excessively, which leads to inferring incorrect matching results for various situations. For example, complex road relationships or movement patterns, such as in urban areas, or serious observation errors and sampling intervals make matching more difficult. In this research, we propose a new algorithm called trendHMM map matching, which complements the assumptions of HMM. This algorithm considers a wider range of dependencies of geopositioning data by incorporating the movements of neighboring data into the matching process. For this purpose, the concept of the window containing adjacent geopositioning data is introduced. Thus trendHMM can utilize relationships among continuous geopositioning data and showed considerable enhancement over HMM-based algorithm. Through experiments, we demonstrated that trendHMM map matching provides more accurate results than the existing HMM map matching for various environments and geopositioning data sets. Our trendHMM algorithm shows up to 17.58% of performance enhancement compared to HMM based one in terms of Route Mismatch Fraction.

Predicting user behavior using data profiling and hidden Markov model

<span lang="EN-US">Mental health disorders affect many aspects of patient’s lives, including emotions, cognition, and especially behaviors. E-health technology helps to collect information wealth in a non-invasive manner, which represents a promising opportunity to construct health behavior markers. Combining such user behavior data can provide a more comprehensive and contextual view than questionnaire data. Due to behavioral data, we can train machine learning models to understand the data pattern and also use prediction algorithms to know the next state of a person’s behavior. The remaining challenges for this issue are how to apply mathematical formulations to textual datasets and find metadata that aids to identify the person’s life pattern and also <span>predict the next state of his comportment. The main idea of this work is to use a hidden Markov model (HMM) to predict user behavior from social media applications by analyzing and detecting states and symbols from the user</span> behavior dataset. To achieve this goal, we need to analyze and detect the states and symbols from the user behavior dataset, then convert the textual data to mathematical and numerical matrices. Finally, apply the HMM model to predict the hidden user behavior states. We tested our program and identified that the log-likelihood was higher and better when the model fits the data. In any case, the results of the study indicated that the program was suitable for the purpose and yielded valuable data.</span>

Solfeggio Teaching Method Based on MIDI Technology in the Background of Digital Music Teaching

This research aims at teaching solfeggio and ear training in college music and proposes a teaching method for college music note recognition that combines the musical instrument digital interface (MIDI) and hidden Markov models (HMM). The experiment showcases that after preprocessing the music frequency sample signal using HMM model, it achieves the target accuracy after 20 times of training. From the HMM transition probability matrix diagram estimated from all training data sets, it can be seen that the transition matrix is close to the diagonal matrix. This indicates its high transfer efficiency. This study compares the HMM model with the other two algorithms, and the results show that its accuracy rate is about 99.56%. The probability of insertion errors and elimination errors is 0.52% and 2.58%. This is superior to the other two algorithms. In summary, the HMM model proposed in the study has extremely strong performance in the teaching of music note feature recognition in universities and can provide better teaching methods.

In-Vehicle Speech Recognition for Voice-Driven UAV Control in a Collaborative Environment of MAV and UAV

Most conventional speech recognition systems have mainly concentrated on voice-driven control of personal user devices such as smartphones. Therefore, a speech recognition system used in a special environment needs to be developed in consideration of the environment. In this study, a speech recognition framework for voice-driven control of unmanned aerial vehicles (UAVs) is proposed in a collaborative environment between manned aerial vehicles (MAVs) and UAVs, where multiple MAVs and UAVs fly together, and pilots on board MAVs control multiple UAVs with their voices. Standard speech recognition systems consist of several modules, including front-end, recognition, and post-processing. Among them, this study focuses on recognition and post-processing modules in terms of in-vehicle speech recognition. In order to stably control UAVs via voice, it is necessary to handle the environmental conditions of the UAVs carefully. First, we define control commands that the MAV pilot delivers to UAVs and construct training data. Next, for the recognition module, we investigate an acoustic model suitable for the characteristics of the UAV control commands and the UAV system with hardware resource constraints. Finally, two approaches are proposed for post-processing: grammar network-based syntax analysis and transaction-based semantic analysis. For evaluation, we developed a speech recognition system in a collaborative simulation environment between a MAV and an UAV and successfully verified the validity of each module. As a result of recognition experiments of connected words consisting of two to five words, the recognition rates of hidden Markov model (HMM) and deep neural network (DNN)-based acoustic models were 98.2% and 98.4%, respectively. However, in terms of computational amount, the HMM model was about 100 times more efficient than DNN. In addition, the relative improvement in error rate with the proposed post-processing was about 65%.

Understanding of the dynamics of mobile reading: An HMM model of user engagement and content consumption

AbstractUnderstanding consumers’ engagement and subsequent content consumption behavior in the mobile context is critical to mobile app providers. In this paper, we develop a Hidden Markov Model (HMM) to capture the dynamics of users’ engagement states and consumption decisions on the number of books/chapters read and the amount of money spent. Our method allows us to simultaneously capture three interdependent usage behaviors using a single integrated model and identify the impact of content loading time and previous reading behavior on users’ engagement dynamics and content consumption. We calibrate the model using a tap stream data set of individual users’ reading activities on a mobile app. Our analysis reveals three distinct engagement states, a low state with inactive users, a medium state with users sampling books, and a high state with users reading intensively. Furthermore, we find that content loading time has higher negative impacts on high‐state users in state transitioning than medium‐state users. In contrast, the days that elapsed since the last visit has a similar negative impact on the users in the high and medium states. The effect of usage frequency on users in state transitioning is always positive. We have also identified the weekend effect and social influence on users’ reading outcomes. Finally, our simulations quantify the shortened content loading time and the days elapsed since the last visit on users’ engagement dynamics and content consumption decisions, which generate important managerial implications for app providers.

A behavioural hierarchical analysis framework in a smart home: Integrating HMM and probabilistic model checking

Smart homes offer great convenience for people living alone and assistance for physically impaired inhabitants. Robust behavioural analysis technology is one of the keys to maximizing the role of the Smart Home. Typically, when it comes to the behavioural analysis of its inhabitants, most researchers have acquired it through data collection from sensors, cameras, and portable Bluetooth sensors. However, a gap in research exists concerning activity recognition in the context of the users physical location in the environment. In this paper, we propose a hierarchical framework based on Hidden Markov Model (HMM) and suggest dividing the behavioural sequence analysis into two layers: spatial transfer and sensor transfer. In addition, we apply probabilistic model checking to verify the properties of each module’s state transfer and obtain the probability of occurrence of the corresponding behavioural sequence. By integrating an implicit Markov model and probabilistic model checking, we effectively analyse the composition and probability of occurrence of three arbitrary sequences of complex behaviours. Finally, anomaly detection and behavioural guidance are discussed based on the proposed behavioural analysis methods.

Effective Epileptic Seizure Detection by Classifying Focal and Non-focal EEG Signals using Human Learning Optimization-based Hidden Markov Model

Electroencephalogram (EEG) recordings are analyzed to make a diagnosis of neurological diseases like epilepsy before surgical intervention. It is essential to examine the EEG signals of individuals with epilepsy to separate between focal and non-focal seizure sources. In this research, a train-optimized hidden Markov model (HMM) based on human learning optimization is proposed to automatically detect the epileptic seizure by distinguishing the focal and non-focal epileptic EEG signals. Human learning optimization (HLO) plays a crucial part in electrode selection, which aids in the subsequent feature extraction process, reduces the model’s complexity, and chooses the features with the best discriminating properties. The HLO-train optimized HMM model receives the feature vector made up of frequency band-based features, which is then tuned to optimum using the proposed model to improve detection performance. The searching and learning skills of the human is included in the development of the HLO algorithm, that helps in boosting the convergence of the model adds value to the research. The Epileptic seizure is effectively recognized using the model and the efficiency of the research is proved by measuring the evaluation metrices. The HLO-train optimized HMM model attained for the CHB-MIT scalp EEG database with an accuracy of 99.158 %, sensitivity of 98.086 %, and specificity of 99.155 %, while the accuracy of 94.231 %, sensitivity of 92.075 %, and specificity of 99.926 % with the Siena scalp dataset, which is efficient, when compared with state of art methods.

A Layered, Hybrid Machine Learning Analytic Workflow for Mouse Risk Assessment Behavior.

Accurate and efficient quantification of animal behavior facilitates the understanding of the brain. An emerging approach within machine learning (ML) field is to combine multiple ML-based algorithms to quantify animal behavior. These so-called hybrid models have emerged because of limitations associated with supervised [e.g., random forest (RF)] and unsupervised [e.g., hidden Markov model (HMM)] ML models. For example, RF models lack temporal information across video frames, and HMM latent states are often difficult to interpret. We sought to develop a hybrid model, and did so in the context of a study of mouse risk assessment behavior. We used DeepLabCut to estimate the positions of mouse body parts. Positional features were calculated using DeepLabCut outputs and were used to train RF and HMM models with equal number of states, separately. The per-frame predictions from RF and HMM models were then passed to a second HMM model layer ("reHMM"). The outputs of the reHMM layer showed improved interpretability over the initial HMM output. Finally, we combined predictions from RF and HMM models with selected positional features to train a third HMM model ("reHMM+"). This reHMM+ layered hybrid model unveiled distinctive temporal and human-interpretable behavioral patterns. We applied this workflow to investigate risk assessment to trimethylthiazoline and snake feces odor, finding unique behavioral patterns to each that were separable from attractive and neutral stimuli. We conclude that this layered, hybrid ML workflow represents a balanced approach for improving the depth and reliability of ML classifiers in chemosensory and other behavioral contexts.

Evaluation of spoken English self-study system considering speech knowledge recognition algorithm

Abstract This paper builds and evaluates a self-learning system for spoken English pronunciation suitable for PC mobile terminal or mobile terminal users, and the core technology of speech recognition is built on Hidden Markov Model (HMM), which is used to decode the speech signal in spoken English learning. This paper studies the related speech recognition theory and signal processing technology, builds a comprehensive English self-learning system in more complex situations and more user types and conducts a comprehensive evaluation of the system. The results show that the overall accuracy of the HMM model in the spoken English recognition and evaluation system built in this paper is good, and the accuracy of the input audio for people of all ages is greater than 90%. In the younger population, the accuracy of male speech signals was the highest in both closed space and open space, reaching 98.12% and 96.53%, respectively. In addition, it is observed that the accuracy gradually decreases as more wrong judgements are made on the speech input signal. When the evaluation is poor, the accuracy of the scoring results decreases to 55%, and when there are fewer voice judgement errors, it is in the excellent range, where the accuracy reached 88%.

Employing a novel hybrid of GA-ANFIS model to predict distribution of whiting fish larvae and juveniles from tropical estuaries in the context of climate change

Global warming may lead to changes in nutritional composition and fish biodiversity in estuaries due to the effect of increasing temperatures on Sillago fish occurrence. Datacollected on whiting fish Sillago from two estuaries in northern Vietnam were used to employ a Genetic algorithm (GA) based Adaptive neuro fuzzy inference system (ANFIS) model for quantifying the relationships between fish distribution and environments. The present study showed that the occurrences of Sillago larvae and juveniles depended on water temperature, salinity and turbidity. The average fish CPUE predicted by GA-ANFIS was consistent with surveyed data, with the coefficient of determination (R2) being 0.63–0.65. Data obtained by applying the Hidden Markov models (HMM) indicated a rapid development of the construction area in the study site, which leads the temperature to increase. Our results suggest that as warming and land-use changes cause estuarine waters to become warmer, climate change may lead to a decline in Sillago fish occurrence numbers and abundance. The study also showed that a temperature of 23 °C is optimal for the development and growth of Sillago fish, which is shown by the higher CPUE of the Ka Long estuary than the Ba Lat estuary. These findings demonstrate applicability and reliability of GA-ANFIS and HMM models for elucidating fish larval and juvenile distribution in tropical estuaries, which will be valuable to inform management and conservation.

An improved discrete harris hawk optimization algorithm for efficient workflow scheduling in multi-fog computing

Workflow scheduling in fog computing is an NP-hard problem that tries to allocate the best possible set of resources for the workflows considering various objectives such as deadlines, costs, energy, and Quality of Service (QoS). However, fog providers may be under heavy loads from the IoT networks, leading to Service Level Agreement (SLA) violations. This article considers an architecture consisting of multiple fog computing providers and provides a Hidden Markov Model (HMM) model for predicting the availability of each fog computing provider regarding factors such as the number of incoming requests to each fog, deadline-missed workflows, and the offloaded tasks from the fogs to the cloud computing. This HMM model is trained using the unsupervised Baum-Welch algorithm, and the availability probability of each fog is computed using the Viterbi algorithm. Then, the fog provider availability probability is used to select a fog computing provider and schedule IoT workflows on it. Additionally, we have improved Harris Hawks Optimization (HHO) algorithm and presented a Discrete Opposition-based of this algorithm, denoted as DO-HHO, for scientific workflow scheduling. The results of extensive experiments conducted using iFogSim demonstrate that our proposed scheme can significantly reduce offloaded tasks on cloud computing, deadline-missed workflows, and SLA violations, outperforming state-of-the-art works.

Hidden markov models in reliability and maintenance

Although the hidden Markov models (HMM) are very popular in many applied areas their use in reliability engineering is limited. Problems such as the selection of the HMM model by choosing the appropriate number of states, or problems of prediction of failures have not been widely covered in the literature. This paper is concerned with the use of HMMs where the state of the system is not directly observable and instead certain indicators of the true situation are provided via a control system. A hidden model can provide key information about the system dependability such as the failed component of the system, the reliability of the system and related measures. A maximum-likelihood estimator of the system reliability is obtained and its asymptotic properties are studied. Finally, the maintenance of the system is considered in this context and new preventive maintenance strategies are defined and their efficiency is measured in terms of expected cost. To prove the finite sample performance of the methodology, an extensive simulation study is developed.

Cloud security based attack detection using transductive learning integrated with Hidden Markov Model

In recent years, organizations and enterprises put huge attention on their network security . The attackers were able to influence vulnerabilities for the configuration of the network through the network. Zero-day (0-day) is defined as vulnerable software or application that is either defined by the vendor or not patched by any vendor of organization. When zero-day attack is identified within the network there is no proper mechanism when observed. To mitigate challenges related to the zero-day attack, this paper presented HMM_TDL, a deep learning model for detection and prevention of attack in the cloud platform. The presented model is carried out in three phases like at first, Hidden Markov Model (HMM) is incorporated for the detection of attacks. With the derived HMM model, hyper alerts are transmitted to the database for attack prevention. In the second stage, a transductive deep learning model with k-medoids clustering is adopted for attack identification. With k-medoids clustering, soft labels are assigned for attack and data and update to the database. In the last phase, with computed HMM_TDL database is updated with computed trust value for attack prevention within the cloud.

Optimal determination of hidden Markov model parameters for fuzzy time series forecasting

This paper presents Fuzzy Time Series (FTS) forecasting technique using Hidden Markov Model (HMM) optimized by Particle Swarm Optimization (PSO) and Genetic algorithm (GA). One of the major limitations of HMM is the lack of efficient method for HMM parameter estimation. Traditional methods like Baum Welch Algorithm (BWA) have been used to address the associated problem with HMM model. The BWA in itself does not truly capture the fuzziness in natural data resulting in the HMM algorithm into local minima. To address this challenge, this paper formulates the HMM parameter estimation problem into an optimization problem optimized by GA and PSO algorithms. To solve the problem of insufficiency in data allied with the HMM model, we adopted a method called smoothing to reduce the occurrence of zero in the observation events. Monte Carlo simulation is used at the end of the forecast to maintain stability and efficiency of the developed approach. The performance of developed model is evaluated using data of daily average temperature and cloud density of Taipei Taiwan, the Internet traffic data of Ahmadu Bello University (ABU) and Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX). Experimental results showed that the proposed forecasting models has a good forecasting accuracy when compared with existing models.