1- Introduction

2- Proposed approach

3- Techniques

4- Results

5- Conclusion and future scope

References

Abstract

Electroencephalography (EEG) is a non-invasive method for the analysis of neurological disorders. Epilepsy is one of the most widespread neurological disorders and often characterized by repeated seizures. This paper intends to conduct an iterative filtering based decomposition of EEG signals to improve upon the accuracy of seizure detection. The proposed approach is evaluated using All India Institute of Medical Science (AIIMS) Patna EEG database and online CHB-MIT surface EEG database. The iterative filtering decomposition technique is applied to extract sub-components from the EEG signal. The feature set obtained from each segmented intrinsic mode function consists of 2-D power spectral density and time-domain features dynamic mode decomposition power, variance, and Katz fractal dimension. The Hidden Markov Model (HMM) based probabilistic model has been designed using the above-stated features representing the seizure and non-seizure EEG events. The EEG signal is classified based on the maximum score obtained from the individual feature-based classifiers. The maximum score derived from each HMM classifier gives the final class information. The proposed decomposition of EEG signals achieved 99.60% and 99.74% accuracy in seizure detection for the online CHB-MIT surface EEG database and AIIMS Patna EEG database, respectively.

Introduction

EEG has a wide application in the diagnosis of neurological diseases such as dementia, migraine, and epilepsy [1]. Epilepsy is a non-communicable disease that can affect people of different ages. Nearly 70 million people are affected by epilepsy all over the world, and in India alone, 12 million people are suffering from epilepsy [2]. The unexpected simultaneous activity of groups of neurons is called a seizure. The duration of seizure in continuous EEG signals varies from 1 to 10 seconds. Depending on the seizure condition, the recording duration varies from minutes to hours. At present, the seizure detection is manual, and accuracy depends largely on the doctor’s experience. There is an urgent requirement to develop an efficient algorithm which can detect seizure efficiently from continuous EEG signal. There are different state-of-the-art methods proposed for seizure detection. Discrete wavelet transform (DWT) and empirical mode decomposition (EMD) are widely used techniques for data decomposition. EMD time-domain features and frequency domain features with a linear discriminant analysis classifier are used for seizure classification [3]. Multivariate empirical mode decomposition, instantaneous rate, instantaneous amplitude, and artificial neural network are efficient in seizure detection [4]. EMD intrinsic mode function (IMF) and Higher-order moment features are found to be effective in seizure detection [5]. DWT coefficients are found to be efficient in seizure and non-seizure classification [6]. Energy ratio from wavelet coefficients and ant colony classifier achieved good classification accuracy [7]. Stationary wavelet transform has a good application in seizure detection.

Abstract

In this paper, a hidden Markov model with auto-correlated observations (HMM-AO) is developed to handle the degradation modeling of manufacturing systems. Unlike the standard hidden Markov models (HMMs), the current observation in the HMM-AO model not only depends on the corresponding hidden system state, but also on the previous observations. A novel algorithm using the expectation maximum is presented to estimate the unknown parameters. Furthermore, missing data and noise that accumulate over time are also considered by modifying the proposed model. Then two remaining useful life prediction methods based on the HMM-AO model are developed. Predictive values of more accuracy can be obtained, since the autocorrelation of observations has been considered and the temporal evolution of degradation processes has been described properly. A case study is illustrated to highlight the advantages of HMM-AO and demonstrate the accuracy and efficiency of the prediction methods. Furthermore, an improved maintenance policy is developed based on the results of remaining useful life prediction. Finally, a comparison with a conventional condition-based maintenance policy is provided to prove the performance of this proposed policy.

Introduction

Maintenance policies play important roles in improving the effectiveness of production systems’ operation. In recent years, maintenance strategies combining data-driven reliability models with observed data of condition monitoring has gained more and more attention [1–3]. Diagnosis and prognosis are two important aspects in the framework of maintenance and have been largely researched [4]. The former focuses on the fault detection [5], while the latter devotes to evaluate the current health state of the system as well as to predict its remaining useful life (RUL). An accurate prediction of the RUL could help develop a more economical and effective maintenance policy.

We propose and validate a new measure of earnings quality based on a hidden Markov model. This measure, termed earnings fidelity, captures how faithful earnings signals are in revealing the true economic state of the firm. We estimate the measure using a Markov chain Monte Carlo procedure in a Bayesian hierarchical framework that accommodates cross-sectional heterogeneity. Earnings fidelity is positively associated with the forward earnings response coefficient. It significantly outperforms existing measures of quality in predicting two external indicators of low-quality accounting: restatements and Securities and Exchange Commission comment letters.

Introduction

Earnings reports are affected by both a firm’s fundamental performance and the measurement process as governed by reporting standards, auditing technology, and managerial discretion (e.g., Dechow et al., 1998; Nikolaev, 2017). Prior research has used the statistical properties of earnings (e.g., smoothness, kinks in earnings, and target beating) and regression-based abnormal accrual models to measure the quality of earnings (for a review, see Dechow et al., 2010). In regression-based models, researchers separate the “abnormal” portion of accruals from the “normal” portion related to fundamental performance and define accruals quality based on abnormal accruals (e.g., Dechow and Dichev, 2002). Yet recent studies point out that an important issue with this approach is that proxies for unobservable “true” earnings may confound performance shocks with reporting discretion (e.g., Guay et al., 1996; Ball, 2013; Dichev et al., 2013; Owens et al., 2017). Our approach uses a structural model to separate accounting quality from the process of true earnings, thereby significantly alleviating the concern of having to rely on the noisy proxies. We assume that each firm transitions among states according to a Markov process. The fundamental performance of the firm is its state, which is either Low (L) or High (H). The firm’s state is unobservable; however, in each period, the firm issues an earnings signal, which is either low (l) or high (h). The probability that the firm issues a particular signal depends on the unobservable state, so inferences about the firm’s state can be made from its earnings signals.

Abstract

Due to the effective utilization rate of the radio frequency spectrum, Cognitive Radio Network (CRN) has gained more popularity in the current research field. The spectrum sensing techniques detect the presence of the idle channel and reallocate to the Secondary Users (SUs). However, the existing spectrum sensing and channel estimation approaches incurs delay while searching for the new channels. To reduce the delay and achieve optimal selection of the channel in CRN, this paper proposes a Hidden Markov Model (HMM)-based channel selection framework. The Time-Slot based optimal routing mechanism is introduced to minimize the delay occurred during the channel search and optimize the range of the spectrum band. Therefore, the bandwidth range of the node is estimated, and the channel is allocated to the SU. The proposed framework exhibit better end-to-end throughput, bandwidth-power product and lower running time, energy consumption, and average end-to-end delay when compared to the existing schemes.

Introduction

CRN [1] solves the radio spectrum scarcity problems and maximizing the spectrum utilization efficiency. The Cognitive Radio (CR) is considered as a vital technology for the next generation that offers the capability for the efficient utilization of the spectrum. CRN is a collection of the primary and secondary networks. The main benefit of the CRs is to reduce the exploitation of the unused spectrum resources by using the characteristics of reconfigurability and intelligent sensing. Instead of the spectrum resources, the channel K-factor and bandwidth are considered as the additional communication parameters [2,3]. The cognitive approach is to permit the coexistence of various networks on the same transmission resources. In the overlay approach, it is ensured that the SUs do not produce any detrimental interference to the Primary Users (PUs). The CR system is based on the sensing and reconfigurability characteristics to fulfill the above requirements. The users can either interconnect with each other in a multi-hop manner or contact the Base Station (BS).

Sequential data often emerge in the measurement of time series, e.g. the acoustic features at the time frames of speech signals [1] and kinematic and biological signals in human movements [2]. Hence, sequential pattern recognition is increasingly used in many data analysis fields such as handwriting recognition [3,4], speech recognition [5,6], activity recognition [7,8], and gesture recognition [2]. In this approach, sequences of information are utilized in decision-making [9]. Sequential pattern recognition in time-series information matches the input data with a pattern, both of which being sequences of information. However, inequality in the length of sequences between the input data and the pattern is a typical problem. Studies in 1960s showed that an appropriate pattern recognition tool should be developed to overcome this problem in speech recognition [10]. One of the main solutions to this problem is the “time-wrapping” idea, presented by Vintsyk [11]. He showed how dynamic programming can be used to find the best assignment between two sequences. In 1970s and 1980s, many researchers proposed various models that mostly included modeling of acoustic information in a network with some finite states, deriving the stochastic information of acoustic features and comparing them with the input signal using the dynamic time-wrapping method [12]. Some examples of these works can be seen in [13–16]. The most well-known method in this field is Hidden Markov Model (HMM) [17]. The capabilities of the HMM such as the ability to match with stochastic sequential time series and suitable training algorithms have made it a pervasive tool in speech recognition. Since mid-1990s, the HMM has been a dominant method in handwriting and speech recognition [18]. Besides, many researchers have demonstrated its superiority to short time classifiers such as MLP [19] and SVM [8]. Despite the powerful mathematical modeling of the HMMs, the main criticisms about them fall into two categories: 1) limitations due to the selected mathematical modeling assumptions.