Browsing by Author "Liyanage, U.P."

Now showing 1 - 9 of 9

Chemometric assessment of bioaccumulation and contamination pathways for toxic metals in diet and environment: implications for chronic kidney disease of unknown etiology (CKDu) in Sri Lankan agricultural regions
(Environmental Monitoring and Assessment, 2024) Perera, R.A.; Perera, R.T.; Liyanage, U.P.; Premaratne, J.; Liyanage, J. A.
Prolonged consumption of foods containing toxic metals can elevate the risk of noncommunicable diseases, including chronic kidney disease of uncertain etiology (CKDu). Despite the increasing number of CKDu cases in Maradankulama and Mahakanadrawa Grama Niladhari Divisions (GN) in Sri Lanka, no prior studies have examined the accumulation of heavy metal(loid)s and their potential association with CKDu prevalence. Furthermore, there is an absence of comprehensive analyses using chemometric techniques such as PCA and hierarchical studies regarding CKDu and heavy metal contamination in Sri Lanka. This study aims to provide initial insights into the accumulation and potential pathways of toxic metals in staple foods within local diets and their subsequent presence in the agricultural environment of examined GNs. Cr, Cd, As, and Ni concentrations in analyzed foods were within permissible limits (MPLs), whereas Pb levels exceeded MPLs in rice (Oryza sativa), gotukola (Centella asiatica), lime (Citrus crenatifolia), and inland fish (Etroplus suratensis). High target hazard quotient (THQt) values in polished rice suggest possible health risks with prolonged intake. Hierarchical analysis suggested a common source of Pb accumulation. PCA and hierarchical clustering revealed the intricate connection between As and Cd, with their concurrent clustering in samples suggesting a potential common origin. This indicates that while individual concentrations comply with acceptable standards, the potential synergistic effects of Cd and As accumulation might pose elevated health risks. Further, the gut tissues of inland fish exhibited pronounced metal concentrations and significant (p < 0.05) positive correlations with toxic metals in the tank sediments suggesting a diet-based bioaccumulation pathway through sediments.
COMPARISON OF PERFORMANCES OFSELECTED FORECASTING MODELS:AN APPLICATION TO DENGUE DATA IN COLOMBO, SRI LANKA
(Department of Statistics & Computer Science, Faculty of Science,& Research & Development Centre for Mathematical Modelling, Faculty of Science, University of Colombo, Sri Lanka., 2021) Attanayake, A.M.C.H.; Perera, S.S.N.; Liyanage, U.P.
Dengue is a one of the diseases in the world which has no exact treatment to recover from the disease. It is rapidly spreading throughout the world by causing large number of deaths [1]. In Sri Lanka, there is an increase of reported dengue cases over recent years. The majority of dengue cases reported in the Colombo district within the Sri Lanka. Effective dengue management and controlling strategies should be implemented to reduce the deaths from the disease. Modelling and predicting the distribution of the dengue will be useful in detecting outbreaks of the dengue and to execute controlling actions beforehand. The objective of this study is to develop an appropriate modelling technique to predict dengue cases. To accomplish this objective, we have chosen our study area as Colombo, Sri Lanka. Seven modelling techniques, namely, Na¨ıve, Seasonal Na¨ıve, Random Walk with Drift, Mean Forecasting, Autoregressive Integrated Moving Average, Exponential Smoothing and TBATS (Trigonometric, Box-Cox Transformation, ARMA errors, Trend and Seasonal components) [2] were chosen in this study to model dengue data. For model development process, monthly reported dengue cases in Colombo from January 2010 to December 2018 were used and validated using the data from January to December in 2019. Mean error, root mean squared error and mean absolute percentage error measurements were used to select the most parsimonious model to predict dengue cases in Colombo, Sri Lanka. Both Exponential and TBATS models were competed in predicting dengue cases by reporting minimum error measures. Therefore, results disclosed that among the selected methods either Exponential Smoothing model or TBATS model can be used to predict dengue cases in Colombo, Sri Lanka.
A Fuzzy Linear Model Using Possibilistic Linear Regression with Least Squares Method: An Application to Dengue and Rainfall Data
(International Postgraduate Research Conference 2019, Faculty of Graduate Studies, University of Kelaniya, Sri Lanka, 2019) Attanayake, A.M.C.H.; Perera, S.S.N.; Liyanage, U.P.
Fuzzy linear models deal with vague and imprecise phenomenon in order to represent better models. These type of models are especially suitable in modelling and predicting dengue disease as the disease associated with various unknown and uncontrollable factors. Further, modelling and predicting the dengue disease is important as it is one of the leading diseases in the world which reports higher number of deaths. This study focuses on modelling reported dengue cases in the Colombo district, Sri Lanka. Particularly, Possibilistic Linear Regression with Least Squares (PLRLS) Method was applied as the modelling procedure. This method was proposed by H. Lee and H. Tanaka in 1999 to deal with crisp inputs and fuzzy output. The rainfall as one of the leading climatic factors that associated with dengue disease included in the model as an independent variable. Data consists of weekly reported dengue cases and weekly average rainfall in the Colombo district from 46th week of 2009 to 12th week of 2015. 2009 to 2014 data were used for model development and rest of the data for model validation. Cross correlation analysis revealed that the rainfall with 10 lags was associated with the reported dengue cases. By considering dengue and rainfall data as crisp inputs, the upper approximation model and lower approximation model were obtained to reflect the fuzziness of the dengue count in the district. The developed coefficients of the fuzzy linear regression were in the form of non-symmetric triangular fuzzy numbers. The left and the right spreads of the central value determined the lower and upper boundary of the interval, respectively, where the corresponding degree of membership equals to 0. The predicted values from the fuzzy regression model and the actual values of the validation set were within the upper and lower approximation models which indicated the possibility of the dengue prediction through PLRLS method. The authors are in the process of testing additional fuzzy linear models by changing fuzzy input/output combinations with incorporating more independent variables.
Modeling and Forecasting the Usage of Cellular and Landline Phones in Sri Lanka
(International Conference on Applied Social Statistics (ICASS) - 2019, Department of Social Statistics, Faculty of Social Sciences, University of Kelaniya, Sri Lanka, 2019) Karunarathne, A.W.S.P.; Perera, M.S.H.; Liyanage, U.P.
Cellular and landline phones are at the forefront of telecommunication. There are two major categories of phones namely cellular phones and landline phones. Landline phones can be classified into another two sub categories as fixed wire landline phone and fixed wireless landline phones. In this study, the total usage of those two categories have considered as the landline phone usage. Since, the mobile segment has instantly developed in Sri Lanka, there can be seen a significant decline in landline phones market.
Modelling and Forecasting the Usage of Cellular and Landline Phones in Sri Lanka: Univariate Time Series Approach
(International Journal of Academic Research, 2020) Karunarathner, A.W.S.P.; Perera, M.S.H.; Liyanage, U.P.
Phones have become a mandatory commodity in human life. Nowadays, there is a very strong increase in the cellular phone market, so we tend to forget landline phone services. According to statistics, cellular phones and landline phones usage up to December 2018 is 32,528,104 and 2,484, 616 respectively. That is, the teledensity (per 100 inhabitants) is 150 for cellular phones and 11.5 for landline phones. Due to the increment of the cellular phones and decrement of the landline phones, it is vitally important to study their behaviour. Therefore, the objective of this paper is to model and forecast the usage of cellular and landline phones in Sri Lanka. The model was developed using 80% of the data and validated with 20%. The usage was modelled with Autoregressive Integrated Moving Average (ARIMA) technique. Several models were fitted and based on the lowest Akaike’s Information Criteria (AIC), ARIMA (1,2,1) and ARIMA (2,2,1) were identified as the best-fitted models with forecasting accuracy measured by Mean Absolute Percentage Error (MAPE) values 1.403 and 0.976 for cellular and landline phones usage respectively, concluding that two ARIMA models have a strong potential for forecasting the usage of cellular and landline phones. This model would be important to those who are with the telecoms market to achieve their business goals.
Portfolio optimization using machine learning techniques: An application on Colombo stock exchange
(Faculty of Science, University of Kelaniya, Sri Lanka, 2021) Nakandala, I. A.; Liyanage, U.P.
Stock market price prediction is a challenging task due to the non-linearity and volatility of the financial data. Thereby, forming a portfolio considering accurately predicted future stock prices becomes an even harder task. As a classical approach, most of the researchers apply statistical techniques as analytical tools in financial time-series data analysis and forecasting. However, due to the dominance of the qualitative factors affecting the financial market and their securities, most of the forecasting and other interpretations have less accuracy. Nevertheless, the recent development of computing algorithms, particularly in the field of data science, gives a better opportunity to develop analytical techniques that accurately handle the high uncertainty and the associated volatility of financial data. In this study, classes of Recurrent Neural Network (RNN) algorithms have been used as data science techniques. In particular, the LSTM (Long-Short Term Memory), a special kind of RNN, is utilized to predict the future stock price returns of the Colombo Stock Exchange (CSE), Sri Lanka. Herein, daily assets prices of 20 companies belonging to the S&P SL20 list, and the list of top 100 ranking companies in Sri Lanka in the year 2020, have been analyzed. In the required forecasting, LSTM has been trained using the daily assets closing prices from 1st of January 2010 to 31st of March 2019. The model accuracy measured by Root Mean Square (RMS) averaged 10%. The formation of the portfolio is based on companies that have the highest stock prices and expected stock returns. As a result of this analysis, 7 companies are selected to form different portfolios. To select a portfolio with the highest return with minimum risk, combinations of 5 companies out of 7, i.e., in total 21 combinations of companies, have been analyzed. In portfolio analysis, Markowitz Model (Mean- Variance Optimization Model), Equal-Weighted Model (EQ) and Monte Carlo Simulation (MCS) have been used. Depending on the selection of companies to the portfolio, the model performances are varied. Thus, the best stock allocation resulting the highest expected return with the minimum risk, given by these three models, is selected as the investment plan. Based on the techniques that have been used, the risk could be controlled in the range of 0.3 to 1.1 values.
Portfolio optimization using machine learning techniques: An application on Colombo stock exchange
(Faculty of Science, University of Kelaniya, Sri Lanka, 2021) Nakandala, I. A.; Liyanage, U.P.
Stock market price prediction is a challenging task due to the non-linearity and volatility of the financial data. Thereby, forming a portfolio considering accurately predicted future stock prices becomes an even harder task. As a classical approach, most of the researchers apply statistical techniques as analytical tools in financial time-series data analysis and forecasting. However, due to the dominance of the qualitative factors affecting the financial market and their securities, most of the forecasting and other interpretations have less accuracy. Nevertheless, the recent development of computing algorithms, particularly in the field of data science, gives a better opportunity to develop analytical techniques that accurately handle the high uncertainty and the associated volatility of financial data. In this study, classes of Recurrent Neural Network (RNN) algorithms have been used as data science techniques. In particular, the LSTM (Long-Short Term Memory), a special kind of RNN, is utilized to predict the future stock price returns of the Colombo Stock Exchange (CSE), Sri Lanka. Herein, daily assets prices of 20 companies belonging to the S&P SL20 list, and the list of top 100 ranking companies in Sri Lanka in the year 2020, have been analyzed. In the required forecasting, LSTM has been trained using the daily assets closing prices from 1st of January 2010 to 31st of March 2019. The model accuracy measured by Root Mean Square (RMS) averaged 10%. The formation of the portfolio is based on companies that have the highest stock prices and expected stock returns. As a result of this analysis, 7 companies are selected to form different portfolios. To select a portfolio with the highest return with minimum risk, combinations of 5 companies out of 7, i.e., in total 21 combinations of companies, have been analyzed. In portfolio analysis, Markowitz Model (Mean- Variance Optimization Model), Equal-Weighted Model (EQ) and Monte Carlo Simulation (MCS) have been used. Depending on the selection of companies to the portfolio, the model performances are varied. Thus, the best stock allocation resulting the highest expected return with the minimum risk, given by these three models, is selected as the investment plan. Based on the techniques that have been used, the risk could be controlled in the range of 0.3 to 1.1 values.
Survey on the health status of the undergraduates of Faculty of Science, University of Kelaniya
(Faculty of Science, University of Kelaniya, Sri Lanka, 2020) Umagiliya, K.S.; Mendis, G.L.L.S.,; Caldera, P.A.D.S.P.,; Deshapriya, H.K.I.L.,; Dewanthi, K.A.T.; Malshika, N.N.D.; Serasinghe, B.K.; Tharika, W.G.D.S.; Weerakoon, H.L.A.,; Samarakoon, T.; Kumara, M.S.M.S.; Liyanage, U.P.
The health status of an undergraduate has a huge impact on their individual aspects such as mental and physical wellbeing, academic performances on the university standards as well as on many key fields of a country. However, a small effort is put in identifying the factors affecting the health status and evaluating them with the objective of improving the health status of undergraduates. This survey presents valid evidence about the habits that determine their health status, physically and mentally. A sample of 384 out of 2203 undergraduates was selected from the faculty of science using a stratified sampling technique for the evaluation, considering the academic levels of the undergraduates. According to the analysis, majority of the undergraduates (63.54%) from the faculty of science are having preferable body mass index (BMI) value, but minority of undergraduates are having obesity. A significant amount of undergraduates having underweight and overweight BMI-categories was also observed. Out of the undergraduates, who were having preferable BMI values, a considerable percentage (51.82%) of undergraduates were observed to be consuming 2-3 liters of water per day. Due to the heavy workload in academics resulting the extra works such as assignments, tutorials, course work etc..., a high percentage of undergraduates were not engaging in physical exercises (59.38%) and sports (66.93%). Swimming was observed as the most popular sport among the undergraduates and it was followed by cricket and football. A higher percentage of undergraduates were observed to participate in sports activities and physical exercises in order to maintain good health and to reduce the anxiety and stress. The analysis highlights that the undergraduates who were having preferable BMI values consume 2-3 liters of water per day, engage in sports and physical activities, maintain good food patterns and have good sleep. Even though the above factors were taken into consideration, there could still exist certain other specific factors that have a significant influence on the health status of an undergraduate. Being healthy is rather a lifestyle that constitutes healthier and wise choices for food and level of water consumption, being positive minded etc. Thus, if the challenge of evaluating and optimizing the health status of undergraduates is achieved; they could make into being more content and positive in every aspect of his or her university performances and peer interactions.
Visualization of Positive Semi Definite Matrices
(19th Conference on Postgraduate Research, International Postgraduate Research Conference 2018, Faculty of Graduate Studies,University of Kelaniya, Sri Lanka, 2018) Ranasinghe, L.P.; Liyanage, U.P.; Perera, S.S.N.
This paper studies on how to identify positive semi definite property of a matrix using a plot. The main difference of positive semi definite matrix and negative semi definite matrix is defined by eigen values. All the eigen values of positive semi definite matrix are non-negative. All the eigen values of negative semi definite matrix are non-positive. This study will help to determine positive semi definite property of a matrix without using matrix calculations and in this research paper, we use positive semi definite matrix, negative semi definite matrix, square and symmetric matrix, non symmetric matrix and non square matrix. 10 by 10 matrices were used for the study except non square matrix. Contour plot was used as a visualization tool. Because of the features of the contour plot, the positive semi definite property of a matrix was identified. The main difference between the contour plot of positive semi definite matrix and contour plot of negative semi definite matrix is location of contour centers. If contour plot represents positive semi definite matrix, then contour centers are all over graph. If contour plot represents negative semi definite matrix, then contour centers lie only in the diagonal. Symmetric property was implied by dividing the contour plot into two equal parts through a line along in the diagonal. If X and Y axis have same ranges in the contour plot, then the contour plot represents square matrix. Therefore, the symmetric property and the square property of matrices were identified from contour plot. If contour centers are all over graph in a contour plot of symmetric and square matrix, then the contour plot represents positive semi definite matrix. We can identify positive semi definite property, symmetric property and square property using contour plot.