Gaussian Mixture and Kernel Density-Based Hybrid Model for Volatility Behavior Extraction From Public Financial Data

Tigani, Smail; Chaibi, Hasna; Saadane, Rachid

doi:10.3390/data4010019

Cited by 5 publications

(3 citation statements)

References 21 publications

(28 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Subsequent clustering uses a sliding window to capture a period of data points and analyze for recurrent patterns (Aghabozorgi et al, 2015;Zolhavarieh et al, 2014). The four main methods of subsequent clustering are: (i) hierarchical (Navarro et al, 1997;Bhattacharjee et al, 2019;Zeitsch, 2019); (ii) partitioning (Madhulatha, 2012;Dolnicar, 2002); (iii) density-based (Campello et al, 2013;Liu & Cao, 2020;Tigani et al, 2019) and; (iv) pattern discovery (Aitken et al, 1995;Ahn et al, 2005;Li et al, 2002). These method have all shown applicability to financial data analysis and portfolio construction.…”

Section: Market States Clusteringmentioning

confidence: 99%

Dynamic Portfolio Optimization with Inverse Covariance Clustering

Wang

Aste

2022

SSRN Journal

View full text Add to dashboard Cite

Section: Market States Clusteringmentioning

confidence: 99%

Dynamic Portfolio Optimization with Inverse Covariance Clustering

Wang

Aste

2022

SSRN Journal

View full text Add to dashboard Cite

“…(2) where i = 1,2,3,…,33 (since n=33) and consequently is the observed value of the response variable for the i th location; the estimated intercept; the estimated regression coefficients for the k th independent variable at the i th location; x ik the observed value of the k th independent variable at the i th location; the number of independent variables (CRP, ferritin, sTfR) in the model; and ε i the i th residual (Edayu and Syerinna, 2018;Fotheringham et al, 2002). A spatial weighting matrix, the Gaussian kernel (Bullmann et al, 2018;Ruzgas and Drulytė, 2013;Tigani et al, 2019) and the bi-square kernel function (Mohammadinia et al, 2017;Nugroho and Slamet, 2018), was used to find the regression coefficient for each location. The optimum bandwidth, i.e.…”

Section: Factors Affecting the Hb Levelmentioning

confidence: 99%

Spatial distribution and identifying biochemical factors affecting haemoglobin levels among women of reproductive age for each province in Indonesia: A geospatial analysis

et al. 2022

View full text Add to dashboard Cite

Anaemia is still a public health problem in Indonesia. The iron supplement program, known as Tablet Tambah Darah (Blood Add Tablet) has not yet produced optimal results. This study aimed to identify the cause of anaemia and the factors that influence it. Biochemical indicator data are haemoglobin (Hb), C-reactive protein (CRP), ferritin and serum transferrin receptor (sTfR) from 9,463 women of reproduction age. Data from the Basic Health Research (Riskesdas) project of 2013 were used for the study. ANOVA as well as global and local regression approaches (classical regression and geo-weighted regression) were used to compare the mean Hb and CRP values between provinces and to determine the factors that influence Hb concentrations. The results showed that the distribution of anaemia in Indonesia is uneven and not always caused by iron deficiency. The lowest Hb mean coupled with the highest iron deficiency was found in Papua, where there are high rates of parasitic infections. In contrast, the highest mean Hb coupled with low iron deficiency, and also low infection rates, was found in North Sulawesi. The Hb concentrations were significantly associated by ferritin, CRP and sTfR and there were varying magnitudes between provinces. Although anaemia is mainly influenced by the iron concentration, CRP, ferritin and sTfR can also affect it through their association with inflammatory reactions. Identification of all causes of anaemia in each province needs to be done in the future, while blanket iron supplementation should be reviewed.

show abstract

“…Subsequent clustering uses a sliding window to capture a period of data points and analyze for recurrent patterns [64,65]. The four main methods of subsequent clustering are: (i) hierarchical [66,67,68]; (ii) partitioning [69,70]; (iii) density-based [71,72,73]and; (iv) pattern discovery [74,75,76]. These method have all shown applicability to financial data analysis and portfolio construction.…”

Section: Market States Clusteringmentioning

confidence: 99%

Dynamic Portfolio Optimization with Inverse Covariance Clustering

Wang¹,

Aste²

2021

Preprint

View full text Add to dashboard Cite

Market conditions change continuously. However, in portfolio's investment strategies, it is hard to account for this intrinsic non-stationarity. In this paper, we propose to address this issue by using the Inverse Covariance Clustering (ICC) method to identify inherent market states and then integrate such states into a dynamic portfolio optimization process. Extensive experiments across three different markets, NASDAQ, FTSE and HS300, over a period of ten years, demonstrate the advantages of our proposed algorithm, termed Inverse Covariance Clustering-Portfolio Optimization (ICC-PO). The core of the ICC-PO methodology concerns the identification and clustering of market states from the analytics of past data and the forecasting of the future market state. It is therefore agnostic to the specific portfolio optimization method of choice. By applying the same portfolio optimization technique on a ICC temporal cluster, instead of the whole train period, we show that one can generate portfolios with substantially higher Sharpe Ratios, which are statistically more robust and resilient with great reductions in maximum loss in extreme situations. This is shown to be consistent across markets, periods, optimization methods and selection of portfolio assets.

show abstract

Gaussian Mixture and Kernel Density-Based Hybrid Model for Volatility Behavior Extraction From Public Financial Data

Cited by 5 publications

References 21 publications

Dynamic Portfolio Optimization with Inverse Covariance Clustering

Dynamic Portfolio Optimization with Inverse Covariance Clustering

Spatial distribution and identifying biochemical factors affecting haemoglobin levels among women of reproductive age for each province in Indonesia: A geospatial analysis

Dynamic Portfolio Optimization with Inverse Covariance Clustering

Contact Info

Product

Resources

About