Fisher’s z Distribution-Based Mixture Autoregressive Model

Stock forecasting technology is always a popular research topic because accurate forecasts allow profitable investments and social change. We postulate, based on past research, three major drawbacks for using time series in forecasting stock prices as follows: (1) a simple time-series model provides insufficient explanations for inner and external interactions of the stock market; (2) the variables of a time series behave in strict stationarity, but economic time-series are usually in a nonlinear or nonstationary state and (3) the forecasting factors of multivariable time-series are selected based on researcher’s knowledge, and such a method is a “subjective” way to construct a forecasting model. Therefore, this paper proposes a causal time-series model to select forecasting factors and builds a machine learning forecast model. The “Granger causality test” is utilized first in the proposed model to select the critical factors from technical indicators and market indexes; next, a “multilayer perceptron regression (MLPR)” is employed to construct a forecasting model. This paper collected financial data over a 13-year period (from 2003 to 2015) of the Taiwan stock index (TAIEX) as experimental datasets. Furthermore, the root mean square error (RMSE) was used as a performance indicator, and we use five forecasting models as comparison models. The results reveal that the proposed model outperforms the comparison models in forecasting accuracy and performs well for three key indicators. LAG1, S&P500 and DJIA, are critical factors in all 11 of our time sliding windows (T1–T11). We offer these results to investors to aid in their decision-making processes.

Download Full-text

Statistical Stability and Spatial Instability in Mapping Forest Tree Species by Comparing 9 Years of Satellite Image Time Series

Remote Sensing ◽

10.3390/rs11212512 ◽

2019 ◽

Vol 11 (21) ◽

pp. 2512 ◽

Cited By ~ 7

Author(s):

Nicolas Karasiak ◽

Jean-François Dejoux ◽

Mathieu Fauvel ◽

Jérôme Willm ◽

Claude Monteil ◽

...

Keyword(s):

Time Series ◽

Tree Species ◽

Cross Validation ◽

Satellite Image ◽

Support Vector ◽

Validation Data ◽

Spatial Instability ◽

Classification Errors ◽

Leave One Out ◽

The Impact

Mapping forest composition using multiseasonal optical time series remains a challenge. Highly contrasted results are reported from one study to another suggesting that drivers of classification errors are still under-explored. We evaluated the performances of single-year Formosat-2 time series to discriminate tree species in temperate forests in France and investigated how predictions vary statistically and spatially across multiple years. Our objective was to better estimate the impact of spatial autocorrelation in the validation data on measurement accuracy and to understand which drivers in the time series are responsible for classification errors. The experiments were based on 10 Formosat-2 image time series irregularly acquired during the seasonal vegetation cycle from 2006 to 2014. Due to lot of clouds in the year 2006, an alternative 2006 time series using only cloud-free images has been added. Thirteen tree species were classified in each single-year dataset based on the Support Vector Machine (SVM) algorithm. The performances were assessed using a spatial leave-one-out cross validation (SLOO-CV) strategy, thereby guaranteeing full independence of the validation samples, and compared with standard non-spatial leave-one-out cross-validation (LOO-CV). The results show relatively close statistical performances from one year to the next despite the differences between the annual time series. Good agreements between years were observed in monospecific tree plantations of broadleaf species versus high disparity in other forests composed of different species. A strong positive bias in the accuracy assessment (up to 0.4 of Overall Accuracy (OA)) was also found when spatial dependence in the validation data was not removed. Using the SLOO-CV approach, the average OA values per year ranged from 0.48 for 2006 to 0.60 for 2013, which satisfactorily represents the spatial instability of species prediction between years.

Download Full-text

Statistical Stability and Spatial Instability in Mapping Forest Tree Species by Comparing 9 Years of Satellite Image Time Series

10.20944/preprints201910.0039.v1 ◽

2019 ◽

Author(s):

Nicolas Karasiak ◽

Jean-François Dejoux ◽

Mathieu Fauvel ◽

Jérôme Willm ◽

Claude Monteil ◽

...

Keyword(s):

Time Series ◽

Tree Species ◽

Cross Validation ◽

Satellite Image ◽

Forest Composition ◽

Validation Data ◽

Spatial Instability ◽

Classification Errors ◽

Leave One Out ◽

The Impact

Mapping forest composition using multiseasonal optical time series is still challenging. Highly contrasted results are reported from one study to another suggesting that drivers of classification errors are still under-explored. We evaluated the performances of single-year Formosat-2 time series to discriminate tree species in temperate forests in France and investigated how predictions vary statistically and spatially across multiple years. Our objective was to better estimate the impact of spatial autocorrelation in the validation data on measurement accuracy and to understand which drivers in the time series are responsible for classification errors. The experiments were based on ten Formosat-2 image time series irregularly acquired during the seasonal vegetation cycle from 2006 to 2014. Due to lot of clouds in the year 2006, an alternative 2006 time series using only cloud-free images has been added. Thirteen tree species were classified in each single-year dataset based on the SVM algorithm. The performances were assessed using a spatial leave-one-out cross validation (SLOO-CV) strategy, thereby guaranteeing full independence of the validation samples, and compared with standard non-spatial leave-one-out cross-validation (LOO-CV). The results show relatively close statistical performances from one year to the next despite the differences between the annual time series. Good agreements between years were observed in monospecific tree plantations of broadleaf species versus high disparity in other forests composed of different species. A strong positive bias in the accuracy assessment (up to 0.4 of Overall Accuracy) was also found when spatial dependence in the validation data was not removed. Using the SLOO-CV approach, the average OA values per year ranged from 0.48 for 2006 to 0.60 for 2013, which satisfactorily represents the spatial instability of species prediction between years.

Download Full-text

Gaussian Mixture Models for Control of Quasi-Passive Spinal Exoskeletons

Sensors ◽

10.3390/s20092705 ◽

2020 ◽

Vol 20 (9) ◽

pp. 2705 ◽

Cited By ~ 2

Author(s):

Marko Jamšek ◽

Tadej Petrič ◽

Jan Babič

Keyword(s):

Mixture Models ◽

Cross Validation ◽

Gaussian Mixture Models ◽

Gaussian Mixture ◽

Work Related ◽

Promising Tool ◽

Control Scheme ◽

Validation Procedure ◽

Leave One Out ◽

Control Output

Research and development of active and passive exoskeletons for preventing work related injuries has steadily increased in the last decade. Recently, new types of quasi-passive designs have been emerging. These exoskeletons use passive viscoelastic elements, such as springs and dampers, to provide support to the user, while using small actuators only to change the level of support or to disengage the passive elements. Control of such devices is still largely unexplored, especially the algorithms that predict the movement of the user, to take maximum advantage of the passive viscoelastic elements. To address this issue, we developed a new control scheme consisting of Gaussian mixture models (GMM) in combination with a state machine controller to identify and classify the movement of the user as early as possible and thus provide a timely control output for the quasi-passive spinal exoskeleton. In a leave-one-out cross-validation procedure, the overall accuracy for providing support to the user was 86 . 72 ± 0 . 86 % (mean ± s.d.) with a sensitivity and specificity of 97 . 46 ± 2 . 09 % and 83 . 15 ± 0 . 85 % respectively. The results of this study indicate that our approach is a promising tool for the control of quasi-passive spinal exoskeletons.

Download Full-text

SWE modelling: comparison between different approaches applied to Germany

10.5194/dach2022-102 ◽

2021 ◽

Author(s):

Fabiana Castino ◽

Bodo Wichura ◽

Harald Schellander ◽

Michael Winkler

Keyword(s):

Time Series ◽

Snow Depth ◽

Cross Validation ◽

Snow Water Equivalent ◽

Snow Density ◽

Design Standards ◽

Model Based ◽

Meteorological Observations ◽

Snow Model ◽

Leave One Out

<p>The characterization of the snow cover by snow water equivalent (SWE) is fundamental in several environmental applications, e.g., monitoring mountain water resources or defining structural design standards. However, SWE observations are usually rare compared to other snow measurements as snow depth (HS). Therefore, model-based methods have been proposed in past studies for estimating SWE, in particular for short timescales (e.g., daily). In this study, we compare two different approaches for SWE-data modelling. The first approach, based on empirical regression models (ERMs), provides the regional parametrization of the bulk snow density, which can be used to estimate SWE values from HS. In particular, we investigate the performances of four different schemes based on previously developed ERMs of bulk snow density depending on HS, date, elevation, and location. Secondly, we apply the semi-empirical multi-layer &#916;snow model, which estimates SWE solely based on snow depth observations. The open source &#916;snow model has been recently used for deriving a snow load map for Austria, resulting in an improved Austrian standard. A large dataset of HS and SWE observations collected by the National Weather Service in Germany (DWD) is used for calibrating and validating the models. This dataset consists of daily HS and three-times-a-week SWE observations from in total ~1000 stations operated by DWD over the period from 1950 to 2020. A leave-one-out cross validation is applied to evaluate the performance of the different model approaches. It is based on 185 time series of HS and SWE observations that are representative of the diversity of the regional snow climatology of Germany. Cross validation reveals for all ERMs: 90% of the modelled SWE time series have a root mean square error (RMSE) and a bias lower than 45 kg/m&#178; and 2 kg/m&#178;, respectively. The &#916;snow model shows the best performance with 90% of the modelled SWE time series having an RMSE lower than 30 kg/m&#178; and bias similar to the ERMs. This comparative study provides new insights on the reliability of model-based methods for estimating SWE values. The results show that the &#916;snow model and, to a lower degree, the developed ERMs can provide satisfactory performances even on short timescales. This suggest that these models can be used as reliable alternative to more complex thermodynamic snow models, even more if long-term meteorological observations aside HS are scarce.</p>

Download Full-text

Scalable and Generalizable Social Bot Detection through Data Selection

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v34i01.5460 ◽

2020 ◽

Vol 34 (01) ◽

pp. 1096-1103 ◽

Cited By ~ 16

Author(s):

Kai-Cheng Yang ◽

Onur Varol ◽

Pik-Mai Hui ◽

Filippo Menczer

Keyword(s):

Cross Validation ◽

State Of The Art ◽

Rapid Development ◽

Model Performance ◽

Training Data ◽

Model Accuracy ◽

Proposed Model ◽

Information Manipulation ◽

Bot Detection ◽

Development State

Efficient and reliable social bot classification is crucial for detecting information manipulation on social media. Despite rapid development, state-of-the-art bot detection models still face generalization and scalability challenges, which greatly limit their applications. In this paper we propose a framework that uses minimal account metadata, enabling efficient analysis that scales up to handle the full stream of public tweets of Twitter in real time. To ensure model accuracy, we build a rich collection of labeled datasets for training and validation. We deploy a strict validation system so that model performance on unseen datasets is also optimized, in addition to traditional cross-validation. We find that strategically selecting a subset of training data yields better model accuracy and generalization than exhaustively training on all available data. Thanks to the simplicity of the proposed model, its logic can be interpreted to provide insights into social bot characteristics.

Download Full-text

GO-GJRSK Model with Application to Higher Order Risk-Based Portfolio

Mathematics ◽

10.3390/math8111990 ◽

2020 ◽

Vol 8 (11) ◽

pp. 1990

Author(s):

Kei Nakagawa ◽

Yusuke Uchiyama

Keyword(s):

Time Series ◽

Stock Prices ◽

Financial Time Series ◽

Higher Order ◽

Estimation Methods ◽

Time Series Modeling ◽

Higher Order Moments ◽

Financial Time ◽

Proposed Model ◽

Multivariate Financial Time Series

There are three distinguishing features in the financial time series, such as stock prices, are as follows: (1) Non-normality, (2) serial correlation, and (3) leverage effect. All three points need to be taken into account to model the financial time series. However, multivariate financial time series modeling involves a large number of stocks, with many parameters to be estimated. Therefore, there are few examples of multivariate financial time series modeling that explicitly deal with higher-order moments. Furthermore, there is no multivariate financial time series model that takes all three characteristics above into account. In this study, we propose the generalized orthogonal (GO)-Glosten, Jagannathan, and Runkle GARCH (GJR) model which extends the GO-generalized autoregressive conditional heteroscedasticity (GARCH) model and incorporates the three features of the financial time series. We confirm the effectiveness of the proposed model by comparing the performance of risk-based portfolios with higher-order moments. The results show that the performance with our proposed model is superior to that with baseline methods, and indicate that estimation methods are important in risk-based portfolios with higher moments.

Download Full-text

Pattern classification of time-series signals using Fisher kernels and support vector machines

10.32920/ryerson.14649759.v1 ◽

2021 ◽

Author(s):

Yashodhan Rajiv Athavale

Keyword(s):

Time Series ◽

Support Vector Machines ◽

Stock Market ◽

Stock Prices ◽

Binary Classification ◽

Gaussian Mixture ◽

Classification Error ◽

Support Vector ◽

Classification Error Rate ◽

Vector Machines

The objective of this study is to assess the performance and capability of a kernel-based machine learning method for time-series signal classification. Applying various stages of dimension transformation, training, testing and cross-validation, we attempt to perform a binary classification using the time-series signals from each category. This study has been applied to two domains: Financial and Biomedical. The financial domain study involves identifying the possibility of collapse or survival of a company trading in the stock market. For assessing the fate of each company, we collect its real stock market data, which is basically a financial time-series composed of weekly closing stock prices in a common time-series interval. This study has been applied to various economic sectors such as Pharmaceuticals and Biotechnology, Automobiles, Oil & Gas, Water Supply etc. The data has been collected using Thomson’s Datastream software. In the biomedical study we are dealing with knee signals collected using the Vibration arthrometry technique. This study involves using the severity of cartilage degeneration for assessing the possibility omachinf a subject getting affected by Osteoarthritis or undergoing knee replacement surgery at a later stage. This non-invasive diagnostic method can also prove be an alternative to various invasive procedures used for detecting osteoarthritis. For this analysis we have used the vibroarthro-signals for about 38 abnormal and 51 normal knee joint case studies. In both studies we apply Fisher Kernels incorporated with Gaussian Mixture Model (GMM) for dimension transformation into feature space created as a three-dimensional plot for visualization. The transformed data is then trained and tested using support vector machines for performing binary classification. From our experiments we observe that our method fits really well for both the studies with the classification error rate between 10% to 15%.

Download Full-text

Pattern classification of time-series signals using Fisher kernels and support vector machines

10.32920/ryerson.14649759 ◽

2021 ◽

Author(s):

Yashodhan Rajiv Athavale

Keyword(s):

Time Series ◽

Support Vector Machines ◽

Stock Market ◽

Stock Prices ◽

Binary Classification ◽

Gaussian Mixture ◽

Classification Error ◽

Support Vector ◽

Classification Error Rate ◽

Vector Machines

The objective of this study is to assess the performance and capability of a kernel-based machine learning method for time-series signal classification. Applying various stages of dimension transformation, training, testing and cross-validation, we attempt to perform a binary classification using the time-series signals from each category. This study has been applied to two domains: Financial and Biomedical. The financial domain study involves identifying the possibility of collapse or survival of a company trading in the stock market. For assessing the fate of each company, we collect its real stock market data, which is basically a financial time-series composed of weekly closing stock prices in a common time-series interval. This study has been applied to various economic sectors such as Pharmaceuticals and Biotechnology, Automobiles, Oil & Gas, Water Supply etc. The data has been collected using Thomson’s Datastream software. In the biomedical study we are dealing with knee signals collected using the Vibration arthrometry technique. This study involves using the severity of cartilage degeneration for assessing the possibility omachinf a subject getting affected by Osteoarthritis or undergoing knee replacement surgery at a later stage. This non-invasive diagnostic method can also prove be an alternative to various invasive procedures used for detecting osteoarthritis. For this analysis we have used the vibroarthro-signals for about 38 abnormal and 51 normal knee joint case studies. In both studies we apply Fisher Kernels incorporated with Gaussian Mixture Model (GMM) for dimension transformation into feature space created as a three-dimensional plot for visualization. The transformed data is then trained and tested using support vector machines for performing binary classification. From our experiments we observe that our method fits really well for both the studies with the classification error rate between 10% to 15%.

Download Full-text

Accurate pancreas segmentation using multi-level pyramidal pooling residual U-Net with adversarial mechanism

BMC Medical Imaging ◽

10.1186/s12880-021-00694-1 ◽

2021 ◽

Vol 21 (1) ◽

Author(s):

Meiyu Li ◽

Fenghui Lian ◽

Chunyu Wang ◽

Shuxu Guo

Keyword(s):

Cross Validation ◽

Contextual Information ◽

Model Performance ◽

Dice Similarity Coefficient ◽

Abdominal Ct ◽

Proposed Model ◽

Contrast Enhanced ◽

Multi Level ◽

Pancreas Segmentation ◽

Fold Cross Validation

Abstract Background A novel multi-level pyramidal pooling residual U-Net with adversarial mechanism was proposed for organ segmentation from medical imaging, and was conducted on the challenging NIH Pancreas-CT dataset. Methods The 82 pancreatic contrast-enhanced abdominal CT volumes were split via four-fold cross validation to test the model performance. In order to achieve accurate segmentation, we firstly involved residual learning into an adversarial U-Net to achieve a better gradient information flow for improving segmentation performance. Then, we introduced a multi-level pyramidal pooling module (MLPP), where a novel pyramidal pooling was involved to gather contextual information for segmentation, then four groups of structures consisted of a different number of pyramidal pooling blocks were proposed to search for the structure with the optimal performance, and two types of pooling blocks were applied in the experimental section to further assess the robustness of MLPP for pancreas segmentation. For evaluation, Dice similarity coefficient (DSC) and recall were used as the metrics in this work. Results The proposed method preceded the baseline network 5.30% and 6.16% on metrics DSC and recall, and achieved competitive results compared with the-state-of-art methods. Conclusions Our algorithm showed great segmentation performance even on the particularly challenging pancreas dataset, this indicates that the proposed model is a satisfactory and promising segmentor.

Download Full-text