Kasun Bandara
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Kasun Bandara has a total of 194 co-authors. Here are the top co-authors by number of shared publications:
- Julian de Hoog
Department of Mechanical Engineering School of Electrical, Mechanical and Infrastructure Engineering The University of Melbourne Melbourne Australia
7 shared publications - Maneesha Perera
Department of Mechanical Engineering School of Electrical, Mechanical and Infrastructure Engineering The University of Melbourne Melbourne Australia
7 shared publications Christoph Bergmeir
6 shared publications- Christoph Bergmeir
Department of Computer Science and Artificial Intelligence, University of Granada (UGR), Av. del Conocimiento, 37, 18016 Granada, Spain
6 shared publications - Saman Kumara Halgamuge
The University of Melbourne
6 shared publications - Rakshitha Godahewa
Department of Data Science and Artificial Intelligence, Monash University, Melbourne, VIC, Australia
4 shared publications
Latest Publications
- Open Access
- 9 May 2026
- Published by Elsevier in SSRN Electronic Journal
Distributed solar generation forecasting using attention-based deep neural networks for cloud movement prediction
Accurate forecasts of distributed solar generation are necessary to maintain grid stability amid the increased uptake of distributed solar photovoltaic (PV) systems. However, the high variability of solar generation over short time intervals (seconds to minutes) caused by cloud movement makes this forecasting task difficult. To address this, using cloud images, which capture the second-to-second changes in cloud cover affecting solar generation, has shown promise. Recently, deep neural networks with attention that focus on important regions of an image have been applied with success in many computer vision applications. However, their use for forecasting cloud movement has not yet been extensively explored. In this work, we propose a novel attention-enhanced convolutional long short-term memory network to forecast cloud movement using infrared satellite imagery and investigate how improvements in cloud movement prediction translate into improved short-term forecasting of distributed solar generation. We evaluate the impact of attention in cloud movement forecasting by comparing attention-based and non-attention-based approaches. In addition, we provide insights into differences in solar forecast performance under cloud conditions corresponding to high and low cloud altitudes inferred from infrared imagery. We find that for clouds at high altitudes, the cloud predictions obtained using attention-based methods result in solar forecast skill score improvements of 5.86% or more compared to non-attention-based methods.
- 9 May 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDistributed solar generation forecasting using attention-based deep neural networks for cloud movement prediction
Accurate forecasts of distributed solar generation are necessary to maintain grid stability amid the increased uptake of distributed solar photovoltaic (PV) systems. However, the high variability of solar generation over short time intervals (seconds to minutes) caused by cloud movement makes this forecasting task difficult. To address this, using cloud images, which capture the second-to-second changes in cloud cover affecting solar generation, has shown promise. Recently, deep neural networks with attention that focus on important regions of an image have been applied with success in many computer vision applications. However, their use for forecasting cloud movement has not yet been extensively explored. In this work, we propose a novel attention-enhanced convolutional long short-term memory network to forecast cloud movement using infrared satellite imagery and investigate how improvements in cloud movement prediction translate into improved short-term forecasting of distributed solar generation. We evaluate the impact of attention in cloud movement forecasting by comparing attention-based and non-attention-based approaches. In addition, we provide insights into differences in solar forecast performance under cloud conditions corresponding to high and low cloud altitudes inferred from infrared imagery. We find that for clouds at high altitudes, the cloud predictions obtained using attention-based methods result in solar forecast skill score improvements of 5.86% or more compared to non-attention-based methods.
- Open Access
- 14 February 2026
- Published by Elsevier in SSRN Electronic Journal
Distributed solar generation forecasting using attention-based deep neural networks for cloud movement prediction
Accurate forecasts of distributed solar generation are necessary to reduce negative impacts resulting from the increased uptake of distributed solar photovoltaic (PV) systems. However, the high variability of solar generation over short time intervals (seconds to minutes) caused by cloud movement makes this forecasting task difficult. To address this, using cloud images, which capture the second-to-second changes in cloud cover affecting solar generation, has shown promise. Recently, deep neural networks with attention that focus on important regions of an image have been applied with success in many computer vision applications. However, their use for forecasting cloud movement has not yet been extensively explored. In this work, we propose a novel attention-enhanced convolutional long short-term memory network to forecast cloud movement using infrared satellite imagery and investigate how improvements in cloud movement prediction translate into improved short-term forecasting of distributed solar generation. We evaluate the impact of attention in cloud movement forecasting by comparing attention-based and non-attention-based approaches. In addition, we provide insights into differences in solar forecast performance under cloud conditions corresponding to high and low cloud altitudes inferred from infrared imagery. We find that for clouds at high altitudes, the cloud predictions obtained using attention-based methods result in solar forecast skill score improvements of 5.86% or more compared to non-attention-based methods.
- 14 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDistributed solar generation forecasting using attention-based deep neural networks for cloud movement prediction
Accurate forecasts of distributed solar generation are necessary to reduce negative impacts resulting from the increased uptake of distributed solar photovoltaic (PV) systems. However, the high variability of solar generation over short time intervals (seconds to minutes) caused by cloud movement makes this forecasting task difficult. To address this, using cloud images, which capture the second-to-second changes in cloud cover affecting solar generation, has shown promise. Recently, deep neural networks with attention that focus on important regions of an image have been applied with success in many computer vision applications. However, their use for forecasting cloud movement has not yet been extensively explored. In this work, we propose a novel attention-enhanced convolutional long short-term memory network to forecast cloud movement using infrared satellite imagery and investigate how improvements in cloud movement prediction translate into improved short-term forecasting of distributed solar generation. We evaluate the impact of attention in cloud movement forecasting by comparing attention-based and non-attention-based approaches. In addition, we provide insights into differences in solar forecast performance under cloud conditions corresponding to high and low cloud altitudes inferred from infrared imagery. We find that for clouds at high altitudes, the cloud predictions obtained using attention-based methods result in solar forecast skill score improvements of 5.86% or more compared to non-attention-based methods.
- Open Access
- 8 October 2025
- Published by Wiley in Journal of Forecasting
- Vol. 45 (1) , 293-315
Enhancing Demand Forecasting in Retail: A Comprehensive Analysis of Sales Promotional Effects on the Entire Demand Life Cycle
Sales promotions pose challenges to retail operations by causing sudden fluctuations in demand, not only during the promotional period but also across the entire sales promotional life cycle. Previous research has predominantly focused on promotional and nonpromotional periods, often overlooking the postpromotional phase, where demand decreases due to consumer stockpiling during promotions. To address this research gap, we investigate both traditional statistical forecasting methods and contemporary approaches, such as global models, implemented using gradient boosting and deep learning techniques. We assess their performance throughout the entire demand life cycle. We employ the base‐lift approach as our benchmark model, commonly used in the retail sector. Our study results confirm that machine learning methods effectively manage demand volatility induced by retail promotions while enhancing forecast accuracy across the demand life cycle. The base‐lift model performs comparably to alternative machine learning methods, albeit with the additional effort required for data cleansing. Our proposed forecasting framework possesses the capability to automate the retail forecasting process in the presence of sales promotions, facilitating efficient retail planning. Thus, this research introduces a novel demand forecasting framework that considers the complete demand life cycle for generating forecasts, and we rigorously evaluate it using real‐world data.
- 8 October 2025
- Published by Wiley in Journal of Forecasting
- Vol. 45 (1) , 293-315
Open AccessEnhancing Demand Forecasting in Retail: A Comprehensive Analysis of Sales Promotional Effects on the Entire Demand Life Cycle
Sales promotions pose challenges to retail operations by causing sudden fluctuations in demand, not only during the promotional period but also across the entire sales promotional life cycle. Previous research has predominantly focused on promotional and nonpromotional periods, often overlooking the postpromotional phase, where demand decreases due to consumer stockpiling during promotions. To address this research gap, we investigate both traditional statistical forecasting methods and contemporary approaches, such as global models, implemented using gradient boosting and deep learning techniques. We assess their performance throughout the entire demand life cycle. We employ the base‐lift approach as our benchmark model, commonly used in the retail sector. Our study results confirm that machine learning methods effectively manage demand volatility induced by retail promotions while enhancing forecast accuracy across the demand life cycle. The base‐lift model performs comparably to alternative machine learning methods, albeit with the additional effort required for data cleansing. Our proposed forecasting framework possesses the capability to automate the retail forecasting process in the presence of sales promotions, facilitating efficient retail planning. Thus, this research introduces a novel demand forecasting framework that considers the complete demand life cycle for generating forecasts, and we rigorously evaluate it using real‐world data.
- Open Access
Distributed solar generation forecasting using attention-based deep neural networks for cloud movement prediction
Accurate forecasts of distributed solar generation are necessary to reduce negative impacts resulting from the increased uptake of distributed solar photovoltaic (PV) systems. However, the high variability of solar generation over short time intervals (seconds to minutes) caused by cloud movement makes this forecasting task difficult. To address this, using cloud images, which capture the second-to-second changes in cloud cover affecting solar generation, has shown promise. Recently, deep neural networks with "attention" that focus on important regions of an image have been applied with success in many computer vision applications. However, their use for forecasting cloud movement has not yet been extensively explored. In this work, we propose an attention-based convolutional long short-term memory network to forecast cloud movement and apply an existing self-attention-based method previously proposed for video prediction to forecast cloud movement. We investigate and discuss the impact of cloud forecasts from attention-based methods towards forecasting distributed solar generation, compared to cloud forecasts from non-attention-based methods. We further provide insights into the different solar forecast performances that can be achieved for high and low altitude clouds. We find that for clouds at high altitudes, the cloud predictions obtained using attention-based methods result in solar forecast skill score improvements of 5.86% or more compared to non-attention-based methods.
Open AccessDistributed solar generation forecasting using attention-based deep neural networks for cloud movement prediction
Accurate forecasts of distributed solar generation are necessary to reduce negative impacts resulting from the increased uptake of distributed solar photovoltaic (PV) systems. However, the high variability of solar generation over short time intervals (seconds to minutes) caused by cloud movement makes this forecasting task difficult. To address this, using cloud images, which capture the second-to-second changes in cloud cover affecting solar generation, has shown promise. Recently, deep neural networks with "attention" that focus on important regions of an image have been applied with success in many computer vision applications. However, their use for forecasting cloud movement has not yet been extensively explored. In this work, we propose an attention-based convolutional long short-term memory network to forecast cloud movement and apply an existing self-attention-based method previously proposed for video prediction to forecast cloud movement. We investigate and discuss the impact of cloud forecasts from attention-based methods towards forecasting distributed solar generation, compared to cloud forecasts from non-attention-based methods. We further provide insights into the different solar forecast performances that can be achieved for high and low altitude clouds. We find that for clouds at high altitudes, the cloud predictions obtained using attention-based methods result in solar forecast skill score improvements of 5.86% or more compared to non-attention-based methods.
- Top Cited PapersOpen Access
- 6 March 2024
- Published by Elsevier in Applied Energy
- Vol. 361
Day-ahead regional solar power forecasting with hierarchical temporal convolutional neural networks using historical power generation and weather data
- 6 March 2024
- Published by Elsevier in Applied Energy
- Vol. 361
Top Cited PapersOpen AccessDay-ahead regional solar power forecasting with hierarchical temporal convolutional neural networks using historical power generation and weather data
- Open Access
Day-ahead regional solar power forecasting with hierarchical temporal convolutional neural networks using historical power generation and weather data
Regional solar power forecasting, which involves predicting the total power generation from all rooftop photovoltaic systems in a region holds significant importance for various stakeholders in the energy sector. However, the vast amount of solar power generation and weather time series from geographically dispersed locations that need to be considered in the forecasting process makes accurate regional forecasting challenging. Therefore, previous work has limited the focus to either forecasting a single time series (i.e., aggregated time series) which is the addition of all solar generation time series in a region, disregarding the location-specific weather effects or forecasting solar generation time series of each PV site (i.e., individual time series) independently using location-specific weather data, resulting in a large number of forecasting models. In this work, we propose two deep-learning-based regional forecasting methods that can effectively leverage both types of time series (aggregated and individual) with weather data in a region. We propose two hierarchical temporal convolutional neural network architectures (HTCNN) and two strategies to adapt HTCNNs for regional solar power forecasting. At first, we explore generating a regional forecast using a single HTCNN. Next, we divide the region into multiple sub-regions based on weather information and train separate HTCNNs for each sub-region; the forecasts of each sub-region are then added to generate a regional forecast. The proposed work is evaluated using a large dataset collected over a year from 101 locations across Western Australia to provide a day ahead forecast. We compare our approaches with well-known alternative methods and show that the sub-region HTCNN requires fewer individual networks and achieves a forecast skill score of 40.2% reducing a statistically significant error by 6.5% compared to the best counterpart.
Open AccessDay-ahead regional solar power forecasting with hierarchical temporal convolutional neural networks using historical power generation and weather data
Regional solar power forecasting, which involves predicting the total power generation from all rooftop photovoltaic systems in a region holds significant importance for various stakeholders in the energy sector. However, the vast amount of solar power generation and weather time series from geographically dispersed locations that need to be considered in the forecasting process makes accurate regional forecasting challenging. Therefore, previous work has limited the focus to either forecasting a single time series (i.e., aggregated time series) which is the addition of all solar generation time series in a region, disregarding the location-specific weather effects or forecasting solar generation time series of each PV site (i.e., individual time series) independently using location-specific weather data, resulting in a large number of forecasting models. In this work, we propose two deep-learning-based regional forecasting methods that can effectively leverage both types of time series (aggregated and individual) with weather data in a region. We propose two hierarchical temporal convolutional neural network architectures (HTCNN) and two strategies to adapt HTCNNs for regional solar power forecasting. At first, we explore generating a regional forecast using a single HTCNN. Next, we divide the region into multiple sub-regions based on weather information and train separate HTCNNs for each sub-region; the forecasts of each sub-region are then added to generate a regional forecast. The proposed work is evaluated using a large dataset collected over a year from 101 locations across Western Australia to provide a day ahead forecast. We compare our approaches with well-known alternative methods and show that the sub-region HTCNN requires fewer individual networks and achieves a forecast skill score of 40.2% reducing a statistically significant error by 6.5% compared to the best counterpart.
- 21 September 2023
- Published by Springer Nature
- p. 163-189
Handling Concept Drift in Global Time Series Forecasting
- 21 September 2023
- Published by Springer Nature
- p. 163-189
Handling Concept Drift in Global Time Series Forecasting
- 21 September 2023
- Published by Springer Nature
- p. 107-122
Forecasting with Big Data Using Global Forecasting Models
- 21 September 2023
- Published by Springer Nature
- p. 107-122
Forecasting with Big Data Using Global Forecasting Models
- Open Access
- 25 May 2023
- Published by SAGE Publications in Environment and Planning B: Urban Analytics and City Science
- Vol. 51 (2) , 366-383
Development and evaluation of probabilistic forecasting methods for small area populations
Planning and development decisions in both the government and business sectors often require small area population forecasts. Unfortunately, current methods often produce forecasts that are inaccurate, particularly for remote areas and those with smaller populations. Such inaccuracy necessitates the development and evaluation of methods to forecast and communicate forecast uncertainty, however, little research has been conducted in this domain for small area populations. In this paper, we evaluate a set of probabilistic forecasting methods which include Autoregressive integrated moving average, Exponential Smoothing, THETA, LightGBM and XGBOOST, to produce point forecasts and 80% prediction intervals for Australian SA2 small area populations. We also investigate methods to combine the intervals to produce ensemble forecasts. Our results show that individual probabilistic methods generally produce prediction intervals which underestimate forecast uncertainty. Combining forecasts improves the overall accuracy of point forecasts and the coverage of their intervals, however, coverage still tends to be less than the expected 80% for all but the most conservative combination method.
- 25 May 2023
- Published by SAGE Publications in Environment and Planning B: Urban Analytics and City Science
- Vol. 51 (2) , 366-383
Open AccessDevelopment and evaluation of probabilistic forecasting methods for small area populations
Planning and development decisions in both the government and business sectors often require small area population forecasts. Unfortunately, current methods often produce forecasts that are inaccurate, particularly for remote areas and those with smaller populations. Such inaccuracy necessitates the development and evaluation of methods to forecast and communicate forecast uncertainty, however, little research has been conducted in this domain for small area populations. In this paper, we evaluate a set of probabilistic forecasting methods which include Autoregressive integrated moving average, Exponential Smoothing, THETA, LightGBM and XGBOOST, to produce point forecasts and 80% prediction intervals for Australian SA2 small area populations. We also investigate methods to combine the intervals to produce ensemble forecasts. Our results show that individual probabilistic methods generally produce prediction intervals which underestimate forecast uncertainty. Combining forecasts improves the overall accuracy of point forecasts and the coverage of their intervals, however, coverage still tends to be less than the expected 80% for all but the most conservative combination method.