The performance of machine learning (ML) algorithms depends on the nature of the problem at hand. ML-based modeling, therefore, should employ suitable algorithms where optimum results are desired. The purpose of the current study was to explore the potential applications of ML algorithms in modeling daylight in indoor spaces and ultimately identify the optimum algorithm. We thus developed and compared the performance of four common ML algorithms: generalized linear models, deep neural networks, random forest, and gradient boosting models in predicting the distribution of indoor daylight illuminances. We found that deep neural networks, which showed a determination of coefficient (R²) of 0.99, outperformed the other algorithms. Additionally, we explored the use of long short-term memory to forecast the distribution of daylight at a particular future time. Our results show that long short-term memory is accurate and reliable (R² = 0.92). Our findings provide a basis for discussions on ML algorithms’ use in modeling daylight in indoor spaces, which may ultimately result in efficient tools for estimating daylight performance in the primary stages of building design and daylight control schemes for energy efficiency.

Recently, the importance of green building certification in consideration of environmentally friendly factors is being emphasized more when constructing buildings in South Korea. The Green Standard for Energy and Environmental Design (G-SEED) is one of the strategies used by the Korean government to effectively reduce building environmental loads. However, due to the large investment needed to acquire green building certification, building owners, stakeholders, and designers often contemplate how to balance G-SEED certification benefits and the additional costs they involve. Therefore, the purpose of this study was to assess the benefits of G-SEED certification in terms of post-occupancy financial advantages through a comparative analysis of real estate prices of apartments in the Yeongnam area. All of the major factors affecting apartment real estate prices in South Korea were considered, and the real estate price difference between G-SEED certified and non-certified apartments was determined through a one-sample t-test. The results demonstrated that G-SEED certified apartment real estate prices were 9.52% higher than non-certified apartments. This study concluded that G-SEED certification–related investment is worth the additional cost as it increases the real estate value of a building.

While the literature confirms the existence of a green premium, various researchers have acknowledged the importance of homogeneity in the dataset used for premium assessment. This study developed a systematic approach to extract a sales subsample containing transactions from green-certified apartments and their most similar counterparts based on cluster analysis. The study applied k-means and PAM clustering algorithms to split an extensive sales sample of 81,605 transactions into sales subsamples. A sales subsample containing transactions from green-certified apartments and their peers was extracted and used for green premium estimation. The results indicate that through cluster analysis, a market segment containing above 80% of the total housing transactions from green-certified apartments could be identified. Through the hedonic model, a green premium of 26.2% was identified from the entire sales sample (no market segmentation). However, this value was reduced to 12.2 and 17.8% when estimated from a sales subsample extracted through k-means and PAM clustering, respectively. These findings have implications for the close assessment of green certification impact on the sale prices. In addition, they can also serve as an indicator of housing categories that need extra effort to promote green practices in that particular market segment.

This study evaluated the impact of including a building ceiling depth into a simulation model on skylight efficiency under two climatic conditions (Ulsan and Seoul, South Korea). Using Radiance and EnergyPlus simulation tools integrated in OpenStudio program by National Renewable Energy Laboratory, Golden, Colorado, USA, daylighting and building energy consumption were computed and assessed to evaluate the energy performance and optimization of skylights. Skylight-to-roof ratios from 1% to 25% were analyzed with ceiling depths of 1.5 m to 3 m. The results showed that the range for efficient skylight ratios became smaller with an increase of ceiling depth; in addition, small apertures were more affected by the ceiling depth than were large apertures. Under Ulsan’s climatic conditions, the optimal skylight-to-roof ratios were 8%, 9%, 10%, and 11% for ceiling depths of 1.5 m, 2 m, 2.5 m, and 3 m, respectively. In Seoul, 8% and 9% were the optimum skylight-to-roof ratios for ceiling depths of 1.5 m and 2 m, respectively; no skylight was energy efficient for a ceiling deeper than 2 m. This study indicates that ceiling depth is a critical factor in the evaluation of skylight performance; thus, it should not be excluded from a simulation model, as is often done to simplify simulation modeling.

Urban upgrading programs are implemented to enhance the living conditions in a city. However, the concept of to what extent the life of the targeted communities is improved remains contested and fluid. The purpose of this study was two-fold. First, it evaluated for the first time the indoor thermal conditions in the three main urban housing types in Kigali and analyzed whether the ongoing urban renewal takes into consideration occupant thermal comfort. Second, alternatives to improve thermal performance of urban housing were investigated and recommendations were given. Results indicate that modern urban housing performed significantly greater compared to other housing types, especially during daytime. Indoor thermal conditions in the planned low-income housing were no better—and, in fact, were slightly poorer—than in the informal settlement from which households were to be relocated. Through design parameter optimization, the indoor thermal conditions in the planned low-income housing were improved by 85.5% in the daytime hours and 71.2% in the nighttime hours. These findings are valuable to officials in resettlement programs by providing insights to improve urban dwellers’ living conditions from a wholistic prospective. This can assist related parties in mitigating project failure often caused by local resistance to resettlement.

High-frequency energy data, such as hourly and sub-hourly energy, provide various options for assessing building energy performance. However, the scarcity of such energy data is among the challenges of applying most of the existing energy analysis approaches in large-scale building energy remodeling projects. The purpose of this study is to develop a practical method to define the energy performance of residential house units using monthly energy data that are relatively easy to obtain for existing building stock. In addition, based on the defined energy use characteristics, house units are classified, and energy retrofit measures are proposed for energy-inefficient units. In this study, we applied a change-point regression model to investigate the heterogeneity in the monthly gas consumption of 200 house units sampled from four apartment complexes in Ulsan, Republic of Korea. Using a four-quadrant plane and the fitted model parameters, we identified most energy-inefficient house units and their potential energy-saving measures are assessed. The results indicate that around a 41% energy reduction through enhanced thermal properties and heating systems was achieved. The study responds to the need for a straightforward procedure for identifying and prioritizing the best targets for effective energy upgrades of existing buildings.