Results 1  10
of
22
Cluster analysis of typhoon tracks. Part I: General properties
 J. CLIMATE
, 2007
"... A new probabilistic clustering technique, based on a regression mixture model, is used to describe tropical cyclone trajectories in the western North Pacific. Each component of the mixture model consists of a quadratic regression curve of cyclone position against time. The besttrack 1950–2002 datas ..."
Abstract

Cited by 30 (6 self)
 Add to MetaCart
(Show Context)
A new probabilistic clustering technique, based on a regression mixture model, is used to describe tropical cyclone trajectories in the western North Pacific. Each component of the mixture model consists of a quadratic regression curve of cyclone position against time. The besttrack 1950–2002 dataset is described by seven distinct clusters. These clusters are then analyzed in terms of genesis location, trajectory, landfall, intensity, and seasonality. Both genesis location and trajectory play important roles in defining the clusters. Several distinct types of straightmoving, as well as recurving, trajectories are identified, thus enriching this main distinction found in previous studies. Intensity and seasonality of cyclones, though not used by the clustering algorithm, are both highly stratified from cluster to cluster. Three straightmoving trajectory types have very small withincluster spread, while the recurving types are more diffuse. Tropical cyclone landfalls over East and Southeast Asia are found to be strongly cluster dependent, both in terms of frequency and region of impact. The relationships of each cluster type with the largescale circulation, sea surface temperatures, and the
Joint probabilistic curve clustering and alignment
 In Advances in Neural Information Processing Systems 17
, 2005
"... Clustering and prediction of sets of curves is an important problem in many areas of science and engineering. It is often the case that curves tend to be misaligned from each other in a continuous manner, either in space (across the measurements) or in time. We develop a probabilistic framework that ..."
Abstract

Cited by 26 (0 self)
 Add to MetaCart
(Show Context)
Clustering and prediction of sets of curves is an important problem in many areas of science and engineering. It is often the case that curves tend to be misaligned from each other in a continuous manner, either in space (across the measurements) or in time. We develop a probabilistic framework that allows for joint clustering and continuous alignment of sets of curves in curve space (as opposed to a fixeddimensional featurevector space). The proposed methodology integrates new probabilistic alignment models with modelbased curve clustering algorithms. The probabilistic approach allows for the derivation of consistent EM learning algorithms for the joint clusteringalignment problem. Experimental results are shown for alignment of human growth data, and joint clustering and alignment of gene expression timecourse data. 1
Probabilistic Clustering of Extratropical Cyclones Using Regression Mixture Models
 Climate Dynamics
, 2006
"... A probabilistic clustering technique is developed for classification of wintertime extratropical cyclone (ETC) tracks over the North Atlantic. We use a regression mixture model to describe the longitudetime and latitude–time propagation of the ETCs. A simple tracking algorithm is applied to 6hourl ..."
Abstract

Cited by 20 (5 self)
 Add to MetaCart
A probabilistic clustering technique is developed for classification of wintertime extratropical cyclone (ETC) tracks over the North Atlantic. We use a regression mixture model to describe the longitudetime and latitude–time propagation of the ETCs. A simple tracking algorithm is applied to 6hourly mean sealevel pressure fields to obtain the tracks from either a general circulation model (GCM) or a reanalysis data set. Quadratic curves are found to provide the best description of the data. We select a threecluster classification for both data sets, based on a mix of objective and subjective criteria. The track orientations in each of the clusters are broadly similar for the GCM and reanalyzed data; they are characterized by predominantly southtonorth (S–N), westtoeast (W–E), and southwesttonortheast (SW–NE) tracking cyclones, respectively. The reanalysis cyclone tracks, however, are found to be much more tightly clustered geographically than those of the GCM. For the reanalysis data, a link is found between the occurrence of cyclones belonging to different clusters of trajectoryshape, and the phase of the North Atlantic Oscillation (NAO). The positive
Motion segmentation by modelbased clustering of incomplete trajectories
 In ECML PKDD, volume 6912 of LNAI
, 2011
"... Abstract. In this paper, we present a framework for visual object tracking based on clustering trajectories of image key points extracted from a video. The main contribution of our method is that the trajectories are automatically extracted from the video sequence and they are provided directly to ..."
Abstract

Cited by 5 (3 self)
 Add to MetaCart
(Show Context)
Abstract. In this paper, we present a framework for visual object tracking based on clustering trajectories of image key points extracted from a video. The main contribution of our method is that the trajectories are automatically extracted from the video sequence and they are provided directly to a modelbased clustering approach. In most other methodologies, the latter constitutes a difficult part since the resulting feature trajectories have a short duration, as the key points disappear and reappear due to occlusion, illumination, viewpoint changes and noise. We present here a sparse, translation invariant regression mixture model for clustering trajectories of variable length. The overall scheme is converted into a Maximum A Posteriori approach, where the ExpectationMaximization (EM) algorithm is used for estimating the model parameters. The proposed method detects the different objects in the input image sequence by assigning each trajectory to a cluster, and simultaneously provides the motion of all objects. Numerical results demonstrate the ability of the proposed method to offer more accurate and robust solution in comparison with the mean shift tracker, especially in cases of occlusions.
State Space Models for Seasonal Aggregation in Sales Forecasting
, 2005
"... filter; HoltWinters method This paper describes a way to improve forecasts by simultaneously forecasting a group of products that exhibit a similar seasonal pattern. There have already been several previous publications that demonstrated forecast improvements using seasonal aggregation. However, th ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
filter; HoltWinters method This paper describes a way to improve forecasts by simultaneously forecasting a group of products that exhibit a similar seasonal pattern. There have already been several previous publications that demonstrated forecast improvements using seasonal aggregation. However, these papers focused on various ad hoc methods to combine seasonal indices from aggregated time series. Instead, we develop state space models in which aggregation is naturally incorporated. Our primary contribution is the seasonal aggregation extension of the Harvey’s dummy seasonal model and the trigonometric seasonal model. Using sales data, we show the possible improvement of forecasting accuracy using these aggregation models, compared with forecasting individual time series. The empirical results suggest that the truncated harmonic trigonometric seasonal aggregation model (a trigonometric seasonal model with a reduced number of harmonics) is the most promising approach to perform seasonal aggregation forecasting in terms of forecast accuracy and computational cost. 1
Understanding Crowd Collectivity: A MetaTracking Approach
"... Understanding pedestrian dynamics in crowded scenes is an important problem. Given highly fragmented trajectories as input, we present a novel, fully unsupervised approach to automatically infer the semantic regions in a scene. Once the semantic regions are learned, given a tracklet of a person, o ..."
Abstract
 Add to MetaCart
(Show Context)
Understanding pedestrian dynamics in crowded scenes is an important problem. Given highly fragmented trajectories as input, we present a novel, fully unsupervised approach to automatically infer the semantic regions in a scene. Once the semantic regions are learned, given a tracklet of a person, our model predicts the pedestrian’s starting point and destination. The method is comprised of three steps. First, the spatial domain of the scene is quantized into hexagons and a 2D orientation distribution function (ODF) is learned for each hexagon. A Time Homogenous Markov Chain Metatracking method is used to automatically find the sources and sinks and later find the dominant paths in the scene. In the last step, using a 3term based trajectory clustering method, we predict the source and sink for each pedestrian. Furthermore, we introduce a 2step trajectory reconstruction method to infer the future behavior of each individual in the scene. Qualitative and quantitative experiments on a video surveillance dataset from New York Grand Central Station demonstrate the effectiveness of our method both in finding the semantic regions and grouping of fragmented tracklets. 1.
Nonparametric Bayesian Mixedeffects Models for Multitask Learning
, 2013
"... In many real world problems we are interested in learning multiple tasks while the training set for each task is quite small. When the different tasks are related, one can learn all tasks simultaneously and aim to get improved predictive performance by taking advantage of the common aspects of all t ..."
Abstract
 Add to MetaCart
(Show Context)
In many real world problems we are interested in learning multiple tasks while the training set for each task is quite small. When the different tasks are related, one can learn all tasks simultaneously and aim to get improved predictive performance by taking advantage of the common aspects of all tasks. This general idea is known as multitask learning and it has been successfully investigated in several technical settings, with applications in many areas. In this thesis we explore a Bayesian realization of this idea especially using Gaussian Processes (GP) where sharing the prior and its parameters among the tasks can be seen to implement multitask learning. Our focus is on the functional mixedeffects model. More specifically, we propose a family of novel Nonparametric Bayesian models, Grouped mixedeffects GP models, where each individual task is given by a fixedeffect, taken from one of a set of unknown groups, plus a random individual effect function that captures variations among individuals. The proposed models provide a unified algorithmic framework to solve time series prediction, clustering and classification.
ho, Gre Trajectory
"... clustering Sparse regression modeling fram an i ract. In ies into a maximum a posteriori approach, where the Expectation–Maximization (EM) algorithm is used deran such as ionba traffic bout a to such family which can be successfully applied when the assumptions made about the model of the motion ar ..."
Abstract
 Add to MetaCart
(Show Context)
clustering Sparse regression modeling fram an i ract. In ies into a maximum a posteriori approach, where the Expectation–Maximization (EM) algorithm is used deran such as ionba traffic bout a to such family which can be successfully applied when the assumptions made about the model of the motion are adequately satisfied, including cases of occlusion. Alternatively, particle filters [4], including the condensation algorithm [5], are more general tracking methods without assuming any specific type of densities. this category. The above methods track only one object at a time. Other methods that simultaneously track many objects have also been proposed [11,12]. In [13], multiple objects are tracked by using Graph Cuts [14] over some observations (i.e. possible locations of the object) which are extracted. Another approach is to employ levelsets to represent each object to be tracked [15,16] which may also be useful to handle the case of multiple object tracking [17,18] by optimally grouping regions whose pixels have similar feature signatures. An application in vehicle tracking is presented in [19] where multiple vehicles are tracked by initially assigning q This paper has been recommended for acceptance by Y. Aloimonos. ⇑ Corresponding author.
3654 JOURNAL OF CLIMATE VOLUME 20 Cluster Analysis of Typhoon Tracks. Part II: LargeScale Circulation and ENSO
, 2006
"... A new probabilistic clustering method, based on a regression mixture model, is used to describe tropical cyclone (TC) propagation in the western North Pacific (WNP). Seven clusters were obtained and described ..."
Abstract
 Add to MetaCart
A new probabilistic clustering method, based on a regression mixture model, is used to describe tropical cyclone (TC) propagation in the western North Pacific (WNP). Seven clusters were obtained and described