Sequential Machine Research Articles

Exploring Player Trace Segmentation for Dynamic Play Style Prediction

Existing work on player modeling often assumes that the play style of players is static. However, our recent work shows evidence that players regularly change their play style over time. In this paper we propose a novel player modeling framework to capture this change by using episodic information and sequential machine learning techniques. In particular, we experiment with different trace segmentation strategies for play style prediction. We evaluate this new framework on gameplay data gathered from a game-based interactive learning environment. Our results show that sequential machine learning techniques that incorporate predictions from previous segments outperform non-sequential techniques. Our results also show that too fine (minute-by-minute) or too coarse (whole trace) segmentation of traces decreases performance.

Simulation And Modelling of BCI Based Multi Purpose Wheel Chair For Paralysed People

According to the survey of Reeve foundation and WHO says, there are nearly 1 in 50 people living with paralysis, every year 5.4 million people affected. Paralysis is caused by spinal injury, stroke, multiple sclerosis, cerebral palsy and other causes like motor accidents and victim of violence. It is also called loss of muscle function in some part of the body. In this paper, we systematically survey a different techniques used for brain computer interface and also review the research on non-invasive, electroencephalography (EEG)-based BCI systems for communication and rehabilitation. Main research focus on previous techniques like deep learning or deep neural network, machine learning, Neuro plasticity, support vector machine, artifact suppression and so on. In this Differing from traditional machine learning algorithms, online sequential machine learning algorithm is empowered to learn distinct high-level representations from raw brain signals without manual feature selection. A certain MATLAB program is designed to use these motions. Our results are to exploit the accuracy rate and to generate the assistive devices for restoration of movement and communication strength for physically disabled patients in order to rehabilitate their lost motor abilities.

Feasibility and reliability of sequential logic with gene regulatory networks.

Gene regulatory networks exhibiting Boolean behaviour, e.g. AND, OR or XOR, have been routinely designed for years. However, achieving more sophisticated functions, such as control or computation, usually requires sequential circuits or so-called state machines. For such a circuit, outputs depend both on inputs and the current state of the system. Although it is still possible to design such circuits by analogy with digital electronics, some particularities of biology make the task trickier. The impact of two of them, namely the stochasticity of biological processes and the inhomogeneity in the response of regulation mechanisms, are assessed in this paper. Numerical simulations performed in two use cases point out high risks of malfunctions even for designed GRNs functional from a theoretical point of view. Several solutions to improve reliability of such systems are also discussed.

Delayed Model Matching of Asynchronous Sequential Machines With Discrepancy in the Initial State

Model matching for asynchronous sequential machines (ASMs) is to design a corrective controller so as to match the stable-state behavior of the closed-loop system to that of a reference model. In this paper, we discuss the model matching problem for ASMs with discrepancy in the initial state. Owing to a lot of practical constraints, the controlled ASM may not commence its operation at the same initial state as the model, leading to initial model mismatch. We investigate under which condition an appropriate corrective controller can be designed that solves initial model mismatch. If such a condition is not valid, we address delayed model matching in which the closed-loop system is controlled to be stably equivalent with the model within an allowable delay-bound from the initial operation. FPGA experiments on an asynchronous error counter are provided to show the applicability of the proposed control scheme.

Pattern Recognition of Development Stage of Creepage Discharge of Oil–Paper Insulation under AC–DC Combined Voltage Based on OS-ELM

The recognition of the creepage discharge development process of oil–paper insulation under AC–DC combined voltage is the basis for fault monitoring and diagnosis of converter transformers; however, only a few related studies are available. In this study, the AC–DC combined voltage with a ratio of 1:1 was used to develop a recognition method for the creepage discharge development process of an oil–paper insulation under a cylinder–plate electrode structure. First, the pulse current method was used to collect the discharge signals in the creepage discharge development process. Then, 24 characteristic parameters were extracted from four types of creepage discharge characteristic spectra after dimensionality reduction. Finally, based on the online sequential extreme learning machine (OS-ELM) algorithm, these characteristic parameters were used to recognize the development stage of the creepage discharge of the oil–paper insulation. The results showed that when the size of the sample training set used in the OS-ELM algorithm is close to the number of hidden layer neurons, a high recognition accuracy can be obtained, and the type of activation function has little influence on the recognition accuracy. Four stages of the creepage discharge development process were recognized using the OS-ELM algorithm; the trend was the same as that of the characteristic parameters of the entire creepage discharge development process. The recognition accuracy was 91.4%. The algorithm has a high computing speed and high accuracy and can train data in batches. Therefore, it can be widely used in the field of online monitoring and evaluation of electrical equipment status.

An FPTAS for just-in-time scheduling of a flow shop manufacturing process with different service quality levels

This paper addresses the problem of identifying a profit-maximizing schedule for jobs in a just-in-time production line, where the decision maker can adjust each job’s processing time, and thereby its quality level and the profit that can be gained from it. The system comprises two sequential machines with a transition period between them and a setup time after completion of each job on the second machine. We first construct an exact algorithm that maximizes the profit. Since this problem is NP-hard, we construct a fully polynomial time approximation scheme (FPTAS) to address it and evaluate its computational complexity.

Quantification of different microplastic fibres discharged from textiles in machine wash and tumble drying

Microplastic fibres released in synthetic cloth washing have been shown to be a source of microplastics into the environment. The annual emission of polyester fibres from household washing machines has earlier been estimated to be 150,000 kg in a country with a population of 5.5 × 106 (Finland). The objectives of this study were (1) to quantify the emissions of synthetic textile fibres discharged from five sequential machine washes (fibre number and length) and tumble dryings (fibre mass) and (2) to determine the collection efficiency of two commercial fibre traps. The synthetic fabrics were five types of polyester textiles, one polyamide and one polyacryl. The number of fibres released from the test fabrics in the first wash varied in the range from 1.0 × 105 to 6.3 × 106 kg−1. The fibre lengths showed that the fleece fabrics released, on average, longer fibres than the technical sports t-shirts. The mass of fibres ranged from 10 to 1700 mg/kg w/w in the first drying. Fibre emissions showed a decreasing trend both in sequential washes and dryings. The ratio of the fibre emissions in machine wash to tumble drying varied between the fabrics: the ratio was larger than one to polyester and polyamide technical t-shirts whereas it was much lower to the other tested textiles. GuppyFriend washing bag and Cora Ball trapped 39% and 10% of the polyester fibres discharged in washings, respectively.

Modeling dynamic construction work template from existing scheduling records via sequential machine learning

Today, construction planning and scheduling is almost always performed manually, by experienced practitioners. The knowledge of those individuals is materialized, maintained, and propagated through master schedules and look-ahead plans. While historical project schedules are available, manually mining their embedded knowledge to create generic work templates for future projects or revising look-ahead schedules is very difficult, time-consuming and error-prone. The rigid work templates from prior research are also not scalable to cover the inter and intra-class variability in historical schedule activities. This paper aims at fulfilling these needs via a new method to automatically learn construction knowledge from historical project planning and scheduling records and digitize such knowledge in a flexible and generalizable data schema. Specifically, we present Dynamic Process Templates (DPTs) based on a novel vector representation for construction activities where the sequencing knowledge is modeled with generative Long Short-Term Memory Recurrent Neural Networks (LSTM-RNNs). Our machine learning models are exhaustively tested and validated on a diverse dataset of 32 schedules obtained from real-world projects. The experimental results show our method is capable of learning planning and sequencing knowledge at high accuracy across different projects. The benefits for automated project planning and scheduling, schedule quality control, and automated generation of project look-aheads are discussed in detail.

Robust corrective control against a class of actuator attacks in input/state asynchronous sequential machines

A robust corrective control scheme is proposed for input/state asynchronous sequential machines (ASMs) vulnerable to an attacker that is able to switch actuator commands. We first present mathematical descriptions of the outcome of actuator attacks in the dynamics of ASMs, and the diagnosis procedure of each attack using the state bursts. We then address the existence condition and design algorithm for a corrective controller that overcomes any actuator attack in an asynchronous mechanism. Potential reachability of the controlled ASM is fully utilized to surmount temporary corruption of actuator commands. An illustrative example is provided to show the proposed concept and methodology.

Model matching inclusion for input/state asynchronous sequential machines with constraint on the length of control input sequences

In this paper, the model matching inclusion problem for input/state asynchronous sequential machines (ASMs) is studied in the framework of corrective control. The considered ASM receives the external input with various frequencies which hinder full use of inherent reachability of the machine. We elucidate the existence condition and design procedure for a state feedback corrective controller that makes the closed-loop system have the largest behavior contained by a given reference model. To demonstrate the practicality of the proposed control scheme, experimental studies on an asynchronous error detection and correction (EDAC) module are conducted by implementing the controller and machine on the field-programmable gate array (FPGA) system.

Short-Term Wind Speed Forecasting Based on Ensemble Online Sequential Extreme Learning Machine and Bayesian Optimization

Short-term wind speed forecasting is crucial to the utilization of wind energy, and it has been employed widely in turbine regulation, electricity market clearing, and preload sharing. However, the wind speed has inherent fluctuation, and accurate wind speed prediction is challenging. This paper aims to propose a hybrid forecasting approach of short-term wind speed based on a novel signal processing algorithm, a wrapper-based feature selection method, the state-of-art optimization algorithm, ensemble learning, and an efficient artificial neural network. Variational mode decomposition (VMD) is employed to decompose the original wind time-series into sublayer modes. The binary bat algorithm (BBA) is used to complete the feature selection. Bayesian optimization (BO) fine-tuned online sequential extreme learning machine (OSELM) is proposed to forecast the low-frequency sublayers of VMD. Bagging-based ensemble OSELM is proposed to forecast high-frequency sublayers of VMD. Two experiments were conducted on 10 min datasets from the National Renewable Energy Laboratory (NREL). The performances of the proposed model were compared with various representative models. The experimental results indicate that the proposed model has better accuracy than the comparison models. Among the thirteen models, the proposed VMD-BBA-EnsOSELM model can obtain the best prediction accuracy, and the mean absolute percent error (MAPE) is always less than 0.09.

Network security situation prediction based on combining associated entropy and deep recurrent neural network

AbstractThe existing network security situation prediction algorithm cannot deal with the subjective attitude deviation of multi‐experts, and the traditional sequential machine learning model cannot deal with the problem of deviation accumulation in time period. In this paper, an algorithm of network security situation prediction combining entropy correlation and deep time series network model is proposed. First, the expert fuzzy evaluation index is constructed by trigonometric fuzzy function, and the modified weighted dempster–shafer (DS) evidence reasoning correction index is used, then the loss and possibility matrix features are created, and finally the information security is evaluated by deep time series network. The simulation experiments are carried out on MIT dataset. The experiments analyze whether the features can cope with multi‐expert conflicts, and evaluate the accuracy, robustness, and time efficiency of the algorithm. The experimental results show that the algorithm proposed in this paper has stronger fuzzy evaluation ability, stronger ability to deal with conflict opinions among experts, more accurate prediction of network security situation in time sequence, and higher robustness, but the efficiency of the algorithm has been maintained.

Coordinated corrective control for model matching of asynchronous sequential machines

This paper proposes coordinated corrective control of input/state asynchronous sequential machines (ASMs). The states of the considered ASM are physically or logically separated to which complete observation by a single controller is impossible. A number of local dynamic controllers that have access to local states cooperate to match the stable-state behaviour of the closed-loop system to that of a reference model. We investigate reachability description of ASMs in the configuration of coordinated control and present the existence condition and design procedure for local controllers that solve the model matching problem. An illustrative example is provided to validate the proposed coordinated corrective control scheme.

Spark-based Parallel OS-ELM Algorithm Application for Short-term Load Forecasting for Massive User Data

The data type and quantity of user load data show an exponential growth, so that the traditional load forecasting methods can hardly meet the load forecasting requirements of massive users. Aiming at this problem, a parallel OS-ELM short-term load forecasting model based on Spark is proposed in this article. By analyzing the characteristics of the Spark framework and the MapReduce framework, the Spark big data processing framework is determined as the basic framework for processing massive user load data, and a parallel K-means load clustering model based on Spark is designed. The on-line sequential learning machine OS-ELM makes the hidden layer data of computing each incremental training dataset mutually independent, therefore, a Spark-based parallel OS-ELM (SBPOS-ELM) algorithm is put forward. The proposed model is applied under the smart electricity big data environment and the training samples are selected using the incremental training dataset to make a short-term prediction of the millions of users’ smart meter electricity load, which verifies the feasibility and effectiveness of the proposed model. At last, comparing with other commonly used short-term load forecasting algorithms, the experimental results show that SBPOS-ELM algorithm has higher accuracy and operation efficiency.

Burst-feedback corrective control of asynchronous sequential machines in non-fundamental mode

Corrective control can compensate for the stable-state behavior of asynchronous sequential machines (ASMs) in a desirable way. This paper presents fault tolerant corrective control for a class of ASMs where output burst, a quick succession of output characters produced along asynchronous transitions, is given as output feedback. The considered ASM is vulnerable to transient faults that may occur during transient transitions of the machine, or in non-fundamental mode. Compared with the prior work on those faults occurring in fundamental mode, the control objective and the reachability conditions for fault recovery are more restrictive in the case of non-fundamental mode. Based on corrective control theory, we address a scheme of fault diagnosis and fault tolerant control for the closed-loop system to overcome every transient fault causing unauthorized state transitions. An illustrative example is provided to validate the synthesis procedure of the proposed burst-feedback corrective control scheme.

Model matching of input/state asynchronous sequential machines with actuator saturation and bounded delays

This paper presents model matching of input/state asynchronous sequential machines (ASMs) in which the corrective controller has the restriction of actuator saturation. As this restriction gives rise to a decrease in the length of control input sequences, a tighter existence condition is needed to achieve the control goal. We also propose the delayed model matching problem in association with actuator saturation, where the stable-state behavior of the closed-loop system is made equivalent with that of a reference model within a finite delay. The delay-bound controllability condition is analyzed and the design procedure for an appropriate model matching controller is addressed in the framework of corrective control theory. An illustrative example is provided to demonstrate the proposed notions and controller synthesis.

Static corrective control with model matching indicator and its application to payload data managers

Having no internal states, static corrective controllers are advantageous over dynamic ones in terms of resource usage, albeit requiring stricter existence conditions. This study presents a novel static corrective control strategy that solves the model matching problem of input/state asynchronous sequential machines. Compared with the previous result, the proposed scheme provides a relaxed existence condition by utilising the model matching indicator that specifies the desirable state at hand. It is shown that the proposed static controller can be designed as long as a corresponding dynamic one exists. The improvement of resource usage in controller synthesis is validated by applying the proposed scheme to the payload data manager in the mass memory unit of satellite systems.

A Survey of Machine Learning Approaches for Student Dropout Prediction in Online Courses

The recent diffusion of online education (both MOOCs and e-courses) has led to an increased economic and scientific interest in e-learning environments. As widely documented, online students have a much higher chance of dropping out than those attending conventional classrooms. It is of paramount interest for institutions, students, and faculty members to find more efficient methodologies to mitigate withdrawals. Following the rise of attention on the Student Dropout Prediction (SDP) problem, the literature has witnessed a significant increase in contributions to this subject. In this survey, we present an in-depth analysis of the state-of-the-art literature in the field of SDP, under the central perspective, but not exclusive, of machine learning predictive algorithms. Our main contributions are the following: (i) we propose a comprehensive hierarchical classification of existing literature that follows the workflow of design choices in the SDP; (ii) to facilitate the comparative analysis, we introduce a formal notation to describe in a uniform way the alternative dropout models investigated by the researchers in the field; (iii) we analyse some other relevant aspects to which the literature has given less attention, such as evaluation metrics, gathered data, and privacy concerns; (iv) we pay specific attention to deep sequential machine learning methods—recently proposed by some contributors—which represent one of the most effective solutions in this area. Overall, our survey provides novice readers who address these topics with practical guidance on design choices, as well as directs researchers to the most promising approaches, highlighting current limitations and open challenges in the field.

Efficient static corrective control for model matching and fault tolerance of asynchronous sequential machines

This paper addresses a static corrective control scheme for input/state asynchronous sequential machines (ASMs) with transient faults. While the previous static controllers have tight existence conditions and achieve only model matching, such restraints are relaxed in this study. To this end, a diagnoser identifying model mismatches and transient faults is introduced to facilitate the design of the static controller. Also state bursts, or fast outbursts of state characters generated by the controlled machine during transitions, are used as feedback. In association with the diangoser, we present the existence condition and design algorithm for the static corrective controller that matches the stable-state behavior of the closed-loop system to that of a reference model while invalidating the adverse effect of transient faults. A case study is provided to demonstrate the applicability of the proposed control methodology.

Low Complexity Channel Prediction Using TFOS-ELM Method for Massive MIMO Systems

Multiple-input multiple-output (MIMO) technology can potentially help to achieve high data rates for multiuser communication. To achieve better performance, the channel state information (CSI) is estimated by the pilot. However, the estimated CSI cannot be used in downlinks when the mobile speed is very high, since it becomes outdated due to the rapid channel variation. In a massive MIMO system, the issue of outdated CSI is serious when using traditional techniques. Therefore, in order to obtain accurate CSI, the prediction of future CSI is required. In this paper, a low complexity online extreme learning machine (ELM) is proposed for the online prediction of the fast fading channel. First, we introduce the structure of the online sequential extreme learning machine (OS-ELM) and combine the training process of the OS-ELM with a forgetting mechanism (FM) to predict fast changing channels. Second, we use the truncated polynomial expansion (TPE) to reduce the computational complexity of the OS-ELM with the FM (FOS-ELM). In addition, the performance of the proposed algorithm is verified through simulation results, and we apply channel prediction in the precoding process. It is shown that the communication quality is improved by our channel prediction algorithm.