Probabilistic Tree Research Articles

A wavelet video coding algorithm with balanced significance probability tree based on energy weighting

This work presents a 3-D wavelet video coding algorithm. By analyzing the contribution of each biorthogonal wavelet basis to reconstructed signal’s energy, we weight each wavelet subband according to its basis energy. Based on distribution of weighted coefficients, we further discuss a 3-D wavelet tree structure named balanced significance probability tree, which places the coefficients with similar probabilities of being significant on the same layer. It is implemented by using hybrid spatial orientation tree and temporal-domain block tree. Subsequently, a novel 3-D wavelet video coding algorithm is proposed based on the energy-weighted balanced significance probability tree. Experimental results illustrate that our algorithm always achieves good reconstruction quality for different classes of video sequences. Compared with asymmetric 3-D orientation tree, the average peak signal-to-noise ratio (PSNR) gain of our algorithm are 1.24dB, 2.54dB and 2.57dB for luminance (Y) and chrominance (U,V) components, respectively. Compared with temporal-spatial orientation tree algorithm, our algorithm gains 0.38dB, 2.92dB and 2.39dB higher PSNR separately for Y, U, and V components. In addition, the proposed algorithm requires lower computation cost than those of the above two algorithms.

Forecasting the Prediction or Predict the Forecasting Influenza Inflection — A Precise View of Investigating Butterfly Effects

Influenza infection is a large global issue. In Hong Kong, the common stages of an influenza outbreak are a decrease in temperature (during winter), which leads to the avian flux, then to the animal flux, and gradually infects humans — leading to an outbreak. However, in Australia, the story is quite different. Excessive carbon dioxide emissions cause extreme weather changes and climatic factors such as strong wind, temperature change, and wet weather. The result is an outbreak of human influenza. These factors constitute a series of correlated sequences. The outcome gives rise to a Bayesian decision tree, which predicts the probability of taking vaccine given that the person is infected. This predictive vaccine efficacy is based on Bayes’ probability, rather than the traditional definition of efficacy or effectiveness. This is the definition of “forecasting a prediction”. Conversely, if Bayes’ probability tree is understood, and the same categorised results are grouped by a random variable, the consequences of the domino effect will be formed. More specifically, this “predicts the forecasting”. One of the aforementioned butterfly effect’s applications (besides an influenza outbreak) is the prediction of an earthquake. It can forecast an earthquake together with its corresponding feasible predicting probability and issue necessary warnings in advance, according to daily phenomena. This act has the potential to save millions of lives.

Traffic Modelling of Low Dense Femtocellular Network for Long Term Evolution

One of the features of Long Term Evolution (LTE) is the deployment of femtocells as the underlain cell of a macrocell without any intervention of frequency planning to offload the traffic. In this paper we used Markov chain to derive the expression of blocking probability for both macro and femtocell in terms of traffic parameters of the network. We developed an analytical model to find the expression of probability of forced termination (FT) using combination of mobility model and probability tree considering low dense femtocellular network. Two different trees were designed: a newly originating call which starts its session in a femtocell and that of in a macro cell. The link parameters of small scale fading of wireless network under Multiple Input Multiple Output (MIMO) are combined with the proposed traffic model to get the probability of FT of a real-life network. A new state transition chain was also developed including its solution for LTE traffic of variable bandwidth (BW) and a comparison was made with Erlang’s traffic model.

Cost-comparison analysis of diffusion weighted magnetic resonance imaging (DWMRI) versus second look surgery for the detection of residual and recurrent cholesteatoma

BackgroundCholesteatoma is a destructive, erosive growth of keratinizing squamous epithelium in the middle ear cleft. Following treatment with a canal wall-up (CWU) tympanomastoidectomy, surveillance of residual and recurrent disease has traditionally been achieved through a second look tympanotomy following the initial procedure. Historically, MRI sequences have been inadequate at differentiating between granulation tissue, inflammation, and cholesteatoma. Recent literature has shown diffusion-weighted magnetic resonance imaging (DWMRI) to be a viable alternative to second look surgery for the detection of residual or recurrent disease. The goal of the present study was to perform a cost analysis of DWIMRI versus second look surgery in the detection of residual or recurrent cholesteatoma following combined approach tympanomastoidectomy.MethodsA probabilistic decision tree model was generated from a literature review to compare traditional second look surgery with DWMRI. Cost inputs were obtained from the Ontario Case Costing Initiative, the Ontario Health Insurance Plan (OHIP) schedule of benefits. Costs were reported in Canadian dollars and a payer perspective was adopted. A probabilistic sensitivity analysis was performed.ResultsAccording to the probabilistic sensitivity analysis, mean cost difference of traditional second look tympanotomy versus echo planar imaging (EPI) DWMRI was $180.27CAD, 95%CI [$177.32, $188,32] in favour of second-look tympanotomy. However, mean cost difference of traditional second look tympanotomy versus non-EPI DWMRI was $390.66CAD, 95%CI [$381.52, $399.80] in favour of non-EPI DWMRI.ConclusionsDiffusion-weighted MRI, specifically non-EPI sequences, are a viable cost-saving alternative to second-look tympanotomy in the setting of detecting residual or recurrent cholesteatoma.

Assembly Sequence Planning by Probabilistic Tree Transformation

Various types of computer systems including CAD/CAM systems have been introduced in machine industry. Some of the systems can handle assembly sequence planning, however it requires long time for planning. This paper proposes a method of generating assembly sequences efficiently. This method extracts some parts and/or subassemblies whose possibilities of being removed from a product are strong, and tests whether they can be removed without any geometric interference. By performing these operations repeatedly, the method generates a disassembly sequence of the product, and obtains an assembly sequence by reversing it. The extraction of some parts and/or subassemblies is performed, based on probabilistic tree transformation. The authors present a calculation example by using a software tool integrated with a CAD system.

Constructive logical characterizations of bisimilarity for reactive probabilistic systems

Larsen and Skou characterized bisimilarity over reactive probabilistic systems with a logic including negation, conjunction, and a diamond modality decorated with a probabilistic lower bound. Later on, working on continuous state spaces, Desharnais, Edalat, and Panangaden proved that negation is not necessary to characterize the same equivalence. In this paper, we redemonstrate the former result with a simpler proof and the latter result directly on discrete state spaces without resorting to measure-theoretic arguments. Moreover, we show that conjunction can be replaced by disjunction in both logics, still characterizing the same bisimilarity. To these ends, we introduce reactive probabilistic trees, a fully abstract model for reactive probabilistic systems that allows us to uniformly prove expressiveness of the four probabilistic modal logics by means of a compactness argument. Our proofs are constructive, as they induce for each considered logic an algorithm that builds a distinguishing formula in case of two inequivalent reactive probabilistic systems.

Pheromone emission patterns and courtship sequences across distinct populations within Anastrepha fraterculus (Diptera-Tephritidae) cryptic species complex.

Sexual behavioural isolation can result from sexual selection and represents a relevant factor associated with the speciation process. We analysed the pheromone emission pattern and the courtship of males of five different populations of the Anastrepha fraterculus cryptic complex: Brazil (Vacaria, Tucumán and Piracicaba), Colombia and Peru. The time of pheromone emission was recorded in each population every 30 min during the day. The behavioural sequences of courting were video recorded and analysed using EthoSeq software. Males from different populations have showed different period of pheromone emission - Vacaria, Piracicaba and Tucumán executed calling only during the morning, Colombia only in the afternoon and Peru during both periods. The general frequencies of the courtship units of the males were distinct among the populations. Three groups were formed in the classification from the function of 14 behavioural routines: Vacaria, Piracicaba and Tucumán formed a single group (Brazil-1), while Colombia and Peru formed two distinct groups. In the probabilistic trees generated, the behavioural units that most contributed to the occurrence of copulation were distinct among the three groups formed: Brazil-1 (Contact, Alignment and Arrowhead-1); Colombia (Flying, Mobile, Contact and Alignment); Peru (Flying, Arrowhead-1 and Calling). Our results indicated differences in sexual behaviour that may explain the behavioural isolation found between the distinct groups in addition with the temporal isolation found between the Brazil-1 and Colombia populations. The evolutionary implications for the A. fraterculus cryptic species complex are discussed.

A new coupled disease-awareness spreading model with mass media on multiplex networks

How to understand the propagation properties of infectious diseases among the population is an important topic in the field of epidemic modelling. In this paper, we investigate the interaction between the disease transmission and disease-related awareness spread, and then propose a new coupled disease spreading model on a two-layered multiplex network, where one layer denotes the underlying topology for the epidemics and the other one represents the corresponding topology for the awareness spread. In the current model, once the information about the disease is acquired by some susceptible ones, it is assumed that they will take the preventive measures to reduce the probability of being infected (e.g., social distancing or taking the vaccine). In the epidemic layer, the transmission dynamics among nodes can be characterized with the classic SIR (susceptible-infective-recovered) model, and those individuals who obtain the disease information will cut down their rate of being infected with a multiplicative factor γ during the course of disease transmission; while in the awareness layer, we will employ the UAU (unaware-aware-unaware) model to denote the diffusion of information. Meanwhile, we explore the impact of mass media on the spread of infectious diseases in the model. Based on the microscopic Markovian chain approach, we build the probability tree to describe the switching process among different states and then establish the different equations for the state transition in detail. Furthermore, we analytically derive the epidemic threshold regarding the disease propagation, which is correlated with the multiplex network topology and the coupling relationship between two classes of spreading dynamics, and extensive simulations also validate the theoretical predictions. Current results are beneficial for us to deeply understand the contagion characteristics of real epidemics within the population.

An economic evaluation of management strategies to mitigate the negative effect of twinning in dairy herds

Our objectives were to develop an economic model to estimate the economic impact of twinning in dairy cows and to evaluate management strategies to mitigate the negative economic impact of twinning in dairy herds. A probabilistic tree considering spontaneous embryo reduction, early pregnancy loss, abortion, metritis, retained placenta, and culling rate at 120 d of the second, at the end of the second, and at the end of the third lactation was developed for a single pregnancy; we also developed 3 management options upon diagnosis of a twin pregnancy: (1) do nothing, (2) induce abortion using PGF2α, or (3) attempt manual embryo reduction. A value was given to each branch of the tree by simulating cow states on a farm for 1,400 d to encompass 4 consecutive lactations. The incomes considered in the simulations included milk income over feed cost, income from calves, and slaughter value upon culling. The expenses taken into account depending on each branch included additional inseminations and synchronization protocols, embryo reduction, induction of abortion, replacement heifers, and costs due to metritis and retained placenta. The gross value for a singleton pregnancy and the 3 management options upon diagnosis of a twin pregnancy were calculated by adding the value of all braches multiplied by their probability. The costs for the 3 management options were calculated by subtracting its gross value minus the gross value of a singleton pregnancy. The negative economic impact of a twin pregnancy ranged from $97 to $225 depending on the type of twin pregnancy (unilateral vs. bilateral), parity, and DIM when the twin pregnancy occurred. The overall negative economic impact of twinning on dairy farm profitability in the United States was estimated to be $96 million per year. Attempting manual embryo reduction early during gestation upon diagnosis of a twin pregnancy was the optimal management strategy for mitigating the negative economic impact of twinning under a wide variety of scenarios.

Hierarchical biogeographical processes largely explain the genomic divergence pattern in a species complex of sea anemones (Metridioidea: Sagartiidae: Anthothoe)

The phylogenetic resolution provided by genome-wide data has demonstrated the usefulness of RAD sequencing to tackle long-standing taxonomic questions. Cnidarians have recently become a model group in this regard, yet species delimitation analyses have been mostly performed in octocorals. In this study, we used RAD sequencing to test the species hypotheses in a wide-spread complex of sea anemones (genus Anthothoe), contrasting this new line of evidence with their current classification. The alternative hypotheses were tested using a Bayes Factors delimitation method, and the most probable species tree was then evaluated under different biogeographic scenarios. Our results decisively rejected the current morphology-informed delimitation model and infer the presence of several cryptic species associated with distinct marine ecoregions. This spatial pattern was remarkably consistent throughout the study, highlighting the role of geographic distribution as a powerful explanatory variable of lineages diversification. The southern Gondwana pattern with episodic, jump dispersal events is the biogeographic historical representation that best fits the Anthothoe species tree. The high population differentiation possibly amplified by the occurrence of asexual reproduction makes it difficult to identify genes responsible for local adaptation, however, these seem to be mainly associated with cellular and metabolic processes. We propose a new set of species hypotheses for the Southern Hemispheric Anthothoe clade, based on the pronounced genomic divergence observed among lineages. Although the link between the genetic and phenotypic differentiation remains elusive, newer sequencing technologies are bringing us closer to understanding the evolution of sea anemone diversity and, therefore, how to appropriately classify them.

A likelihood-free Bayesian derivation method for service variants

Application programming interfaces (API), allowing systems to be accessed by the services they expose, have proliferated on the Internet and gained strategic interest in the IT industry. However, integration opportunities for larger, enterprise systems are hampered by complex and overloaded operations of their interfaces, having hundreds of parameters and multiple levels of nesting, corresponding to multiple business entities. Static (code) analysis techniques have been proposed to analyse service interfaces of enterprise systems. They support the derivation of business entities and relationships from the parameters of interface operations, allowing the restructure of operations, based on individual entities. In this paper, we extend the repertoire of static interface analysis to derive service variants, whereby subsets of operation parameters correspond to multiple nested business entity subtypes of variants. Specifically, we apply a Monte Carlo sampling method, based on likelihood-free Bayesian sampling, to traverse large parameter spaces, based on higher probabilistic tree search, to efficiently find subsets of parameters related to prospective subtypes. The results demonstrate a method with significant success rates in massive search spaces, as applied to the FedEx Shipment interface whose operations have in excess of 1000 parameters.

Multi-model ensembles for assessment of flood losses and associated uncertainty

Abstract. Flood loss modelling is a crucial part of risk assessments. However, it is subject to large uncertainty that is often neglected. Most models available in the literature are deterministic, providing only single point estimates of flood loss, and large disparities tend to exist among them. Adopting any one such model in a risk assessment context is likely to lead to inaccurate loss estimates and sub-optimal decision-making. In this paper, we propose the use of multi-model ensembles to address these issues. This approach, which has been applied successfully in other scientific fields, is based on the combination of different model outputs with the aim of improving the skill and usefulness of predictions. We first propose a model rating framework to support ensemble construction, based on a probability tree of model properties, which establishes relative degrees of belief between candidate models. Using 20 flood loss models in two test cases, we then construct numerous multi-model ensembles, based both on the rating framework and on a stochastic method, differing in terms of participating members, ensemble size and model weights. We evaluate the performance of ensemble means, as well as their probabilistic skill and reliability. Our results demonstrate that well-designed multi-model ensembles represent a pragmatic approach to consistently obtain more accurate flood loss estimates and reliable probability distributions of model uncertainty.

Ensemble human movement sequence prediction model with Apriori based Probability Tree Classifier (APTC) and Bagged J48 on Machine learning

The accurate prediction of human movement trajectory has a variety of benefits for many applications such as optimizing nurse’s trajectory in a hospital, optimizing movements of old or disabled people to minimize their routine efforts, etc. To perform human movement prediction, large amount of historical positioning data from sensors has to be collected and mined. We analyzed different human sequential movement prediction approaches and their limitations. In this work, we propose a new classifier named Apriori based Probability Tree Classifier (APTC) which predicts the human movement sequence patterns in indoor environment. The APTC classifier is integrated into Bagged J48 Machine learning algorithm which results in an ensemble model to predict the future human movement patterns. The patterns are mined based on spatial, temporal and social data which add more accuracy to our prediction. Our model also performs clustering mechanism to detect the abnormal patterns.

Estimating panel effects in probabilistic representations of dynamic decision trees using bayesian generalized linear mixture models

When collecting panel data, we need to acknowledge that responses do not represent independent measurements. The known apparatus in transportation research offers several opportunities to estimate panel effects for well-known and widely applied models such as hazard and dynamic logit models. However, the transportation research community is not endowed with a rich set of methods to account for panel effects in dynamic probabilistic decision trees, which have been used as a formalism for the representation of decision heuristics. Building on scarce prior work in statistics, we elaborate an approach to estimate panel effects in dynamic probabilistic decision trees with multinomial action states. Given that panel data naturally have a hierarchical structure with repeated measures nested within individuals, we implement a mixed-effects model that simultaneously accounts for population-level effects (fixed effects), between-individual variances (random effects), and within-individual variances (autocorrelations). The approach uses an iterative estimation procedure between CHAID-based probabilistic tree induction and Bayesian generalized linear mixture modeling (GLMM). When extracting the dynamic probabilistic decision trees, it is assumed that the random effects are known, while it is assumed that the fixed effects are known when estimating the Bayesian GLMM. This iterative process continues until convergence is reached. A Monte Carlo technique is used to navigate between aggregate choice probabilities and individual level multinomial choices. We also test the significance of temporal autocorrelation within individuals. The suggested approach is illustrated using charging station choice of users of Plug-in Electric Vehicles (PEV). Results support the potential value of the suggested approach.

Visualizing the Bayesian 2-test case: The effect of tree diagrams on medical decision making.

In medicine, diagnoses based on medical test results are probabilistic by nature. Unfortunately, cognitive illusions regarding the statistical meaning of test results are well documented among patients, medical students, and even physicians. There are two effective strategies that can foster insight into what is known as Bayesian reasoning situations: (1) translating the statistical information on the prevalence of a disease and the sensitivity and the false-alarm rate of a specific test for that disease from probabilities into natural frequencies, and (2) illustrating the statistical information with tree diagrams, for instance, or with other pictorial representation. So far, such strategies have only been empirically tested in combination for “1-test cases”, where one binary hypothesis (“disease” vs. “no disease”) has to be diagnosed based on one binary test result (“positive” vs. “negative”). However, in reality, often more than one medical test is conducted to derive a diagnosis. In two studies, we examined a total of 388 medical students from the University of Regensburg (Germany) with medical “2-test scenarios”. Each student had to work on two problems: diagnosing breast cancer with mammography and sonography test results, and diagnosing HIV infection with the ELISA and Western Blot tests. In Study 1 (N = 190 participants), we systematically varied the presentation of statistical information (“only textual information” vs. “only tree diagram” vs. “text and tree diagram in combination”), whereas in Study 2 (N = 198 participants), we varied the kinds of tree diagrams (“complete tree” vs. “highlighted tree” vs. “pruned tree”). All versions were implemented in probability format (including probability trees) and in natural frequency format (including frequency trees). We found that natural frequency trees, especially when the question-related branches were highlighted, improved performance, but that none of the corresponding probabilistic visualizations did.

Speech Replay Spoofing Attack Detection System Based on Fusion of Classification Algorithms

Fast development of modern technologies of digital processing and speech recording leads to the fact that it is necessary to take into account the potential threats from the speech replay attacks. We propose our ensemble fusion replay attack detection system. It uses constant Q cepstral coefficients as speech features and short-time mean normalization for their preprocessing. The set of binary classifiers includes multiple Gaussian mixture models based Bayesian classifier, i-vector based Gaussian Probabilistic Linear Discriminant Analysis and XGBoost tree boosting algorithm. Fusion of scores was made by modified logistic regression algorithm from BOSARIS toolbox. ASV Spoof 2017 corpus is utilized in the experiments as the main database for anti-spoofing systems evaluation. Obtained results demonstrate that the proposed system can provide substantially better performance than the baseline Gaussian mixture model classifier. The pre-processing of cepstral features is crucial for the better performance of the system. High evaluation performance can be obtained using only few algorithms in a set. The attained value of equal error rate EER=12.44% for our fusion classifier is competitive with the best results obtained during last two years.DOI 10.14258/izvasu(2018)1-19

Interplay between SIR-based disease spreading and awareness diffusion on multiplex networks

In this paper, we propose a coupled multiplex network framework to model the epidemic spreading and its corresponding information diffusion among a population. In the model, as far as the information perception on the epidemics is concerned, the individuals can be divided into two classes, namely aware or unaware ones; Meanwhile, the awareness diffusion is depicted by utilizing the traditional contact process. From the perspective of infectious disease spreading, the contagion dynamics among nodes can be characterized with the classic SIR (susceptible–infective–recovered) model. Based on the microscopic Markov chain approach, we build the probability tree to describe the switching process between different states, and then intensively perform the theoretical analysis for the state transition. In particular, we analytically derive the epidemic threshold regarding the disease propagation, which is correlated with the multiplex network topology and the coupling relationship between two transmission dynamics. After being compared with extensive numerical Monte Carlo (MC) simulations, it is clearly found that the achieved analytical results concur with the MC simulations. Current results will be beneficial to substantially enhance the predictability of the epidemic outbreaks within many realistic dissemination cases.

Computing and maximizing the exact reliability of wireless backhaul networks

The reliability of a fixed wireless backhaul network is the probability that the network can meet all the communication requirements considering the uncertainty (e.g., due to weather) in the maximum capacity of each link. We provide an algorithm to compute the exact reliability of a backhaul network, given a discrete probability distribution on the possible capacities available at each link. The algorithm computes a conditional probability tree, where at each leaf in the tree a valid routing for the network is evaluated. Any such tree provides bounds on the reliability, and the algorithm improves these bounds by branching in the tree. We also consider the problem of determining the topology and configuration of a backhaul network that maximizes reliability subject to a limited budget. We provide an algorithm that exploits properties of the conditional probability tree used to calculate reliability of a given network design, and we evaluate its computational efficiency.

An Adaptive Quad-Tree Depth Range Prediction Mechanism for HEVC

A next-generation video coding standard High Efficiency Video Coding (HEVC) provides higher video quality and lower compression bit rate but leads to very high encoding complexity, especially in the quad-tree-based coding tree unit partitioning process. To reduce the computational complexity of HEVC, in this paper, we propose an adaptive quad-tree depth range prediction mechanism. First, the proposed mechanism defines the similar region flag to distinguish between the similar region and the non-similar region. Then, two algorithms, the similar region depth range prediction algorithm and the non-similar region depth range prediction algorithm, are proposed. The similar region depth range prediction algorithm estimates the features of the similar region based on the coding unit depth of this region. The optimal depth of this region can be predicted. The non-similar region depth range prediction algorithm can skip low probability tree nodes based on the depth correlation coefficient, which is calculated based on scene content change. Both the similar region depth range prediction algorithm and the non-similar region depth range prediction algorithm show more than 90% predictive accuracy. Experimental results show that under random access configuration and low delay configuration, the proposed mechanism can yield 28.17% and 32.99% computational complexity reduction with negligible rate distortion performance loss, respectively, compared with HM16.9. The results show that the proposed mechanism is expected to be applied in real-time environments.

Image structure-based saliency detection

We consider image intrinsic structure in two aspects: foreground saliency computation and background saliency calculation. On the one hand, a segmented image is represented as a weighted undirected graph, and we calculate the foreground saliency by choosing some superpixels as ranking queries, where we apply a robust background measure to select those for accuracy. On the other hand, we compute the posterior probabilities to measure the background saliency and in terms of the probability, we construct a probability tree via multimerging on superpixels, and then apply the optimization strategy to the background saliency. We evaluate our proposed algorithm on two benchmark datasets and our algorithm yields the competitive results when compared with nine state-of-the-art algorithms in terms of five evaluation metrics.