Return on equity (ROE) and return on assets (ROA) are important indicators that reveal the sustainability of a company’s profitability performance for both managers and investors. The correct prediction of these indicators will provide a basis for the strategic decisions made by the company managers. The estimation of these signs is a significant factor in supporting the decisions and up-to-date knowledge of potential investors. In this study, return on equity and return on assets were estimated using artificial neural networks (ANNs), multiple linear regression (MLR), and support vector regression (SVR) on the financial data of thirteen companies operating in the iron and steel sector. The success of predicting ROA in the designed model was 86.4% for ANN, 79.9% for SVR, and 74% for MLR. The success of estimating the ROE of the same model was 85.8% for ANN, 80.9% for SVR, and 63.8% for MLR. It is concluded that ANN and SVR can produce successful prediction results for ROA and ROE both accurately and reasonably.

Climate awareness caused by the threat of global warming is the number one agenda item for developed and developing economies. Plans developed in this context, environmentally friendly trends in economic activities, and countries’ efforts to adapt to sustainable development have enabled new road maps. The most important of these efforts is the Paris Climate Agreement signed in 2015 and the Green Deal implemented by the European Union (EU) within the framework of this agreement. In this study, the carbon emissions of Turkey, which has important trade relations with the EU, were estimated using machine learning techniques, and a prediction was made for 2030 based on the results obtained. These results were evaluated in line with the targets of the Green Deal. The R² of support vector regression (SVR), which was applied in the model as one of the machine learning techniques, was found to be 98.4%, and it was found to have the highest predictive power. This technique was followed by multiple linear regression (MLR) with a 97.6% success rate and artificial neural networks (ANN) with a 95.8% success rate, respectively. According to the estimates achieved with the most successful model, SVR, Turkey’s carbon emissions are expected to be 723.97 million metric tons (mt) of carbon dioxide (CO₂) in 2030, the target year set by the EU. This level is 42% higher than the target that needs to be achieved given the data existing in 2019. According to these results, Turkey will not be able to reach the targets set by the EU unless it increases its coal-based energy consumption and provides incentives for renewable energy sources.

IT is recognized as the engine of the digital world. The fact that this technology has multiple sub-sectors makes it the driving force of the economy. With these characteristics, the sector is becoming the center of attention of investors. Considering that investors prioritize profitability, it becomes a top priority for managers to make accurate and reliable profitability forecasts. The aim of this study is to estimate the profitability of IT sector firms traded in Borsa Istanbul using machine learning methods. In this study, the financial data of 13 technology firms listed in the Borsa Istanbul Technology index and operating between March 2000 and December 2023 were used. Return on assets (ROA) and return on equity (ROE) were estimated using machine learning methods such as neural networks, multiple linear regression and decision tree regression. The results obtained reveal that the performance of artificial neural networks (ANN) and multiple linear regression (MLR) are particularly effective.

In the literature, there are two basic approaches regarding the determination of house prices. One of them is the prediction of house price using macroeconomic variables in the country where the house is produced, and another one is the price prediction models, which we can express as micro-variables, by considering the features of the house. In this study, the price of the house was attempted to be predicted using machine learning methods by establishing a model with micro variables that reveal the features of the house. The study was conducted in Turkey’ Antalya province, where household housing demand of foreigners is also high. The house advertisements in locations belonging to the lower, middle- and upper-income groups were selected as the sample. In the results, it was observed that the artificial neural network (ANN) method made predictions with more meaningful results compared to support vector regression (SVR) and multiple linear regression (MLR). These results appear to be a viable model for institutions that supply housing, mediate housing sales, and provide housing financing and valuation. It is considered that this model, which can be used to predict fluctuating house prices, especially in developing countries, will regulate the housing market.

In this study, decision tree regression, artificial neural networks (ANN) and support vector machines (SVM) methods are applied by using monthly data for the period 2013-2020 in the estimation of housing sales in Turkey. In the analysis, the volume of individual mortgage loans offered by banks, the average annual interest rate of mortgage loans from macroeconomic and market variables, the consumer price index (CPI), the BIST 100 index, the benchmark bond interest rate, gold prices and the values of the US dollar and Euro Turkish lira and the housing sales price per square meter in Turkey are used. As a result of the analysis carried out on the model created house sales prices in the Turkish housing market have been successfully estimated and in the light of these estimates, it is determined that banks can guide banks in the creation of various credit packages and appropriate loan targets to support the housing sector.

Bu çalışmada Türkiye’de faaliyet gösteren ve atak portföy yapısı ile tanımlanan 12 bireysel emeklilik fonunun Aralık 2005- Ocak 2020 dönemi arasında aylık olarak net varlık değerleri (NAV) çok katmanlı algılayıcı (ÇKA) ve çoklu doğrusal regresyon yöntemleriyle tahmin edilmeye çalışılmıştır. Bunun için 12 bağımsız değişken ve bir bağımlı değişkenden oluşan modeller oluşturulmuş; elde edilen sonuçların başarı oranları ve hata değerleri karşılaştırılmıştır. ÇKA yönteminde giriş katmanında 12 düğüm, 2 gizli katman ve her gizli katmanda 5 düğüm ve çıkış katmanında bir düğüm olan bir ağ modeli tasarlanmıştır. Tasarlanan bu ağ modelinde gizli katma sayısı ve her gizli katmandaki nöron sayısı aynıdır. Çoklu doğrusal regresyon modelinde 12 bağımsız değişken ile bir bağımlı değişken arasındaki bağıntıyı ortaya koymak ve tahminde bulunmak için her bir bireysel emeklilik yatırım fonunun net varlık değeri için bir model oluşturulmuştur. ÇKA ağının başarı oranı %77,40, R2 0,986, RMSE 0,011, MAE 0,006, MAPE 0,076; ÇDR ise başarı oranı ise %44,54, R2 0,981, RMSE 0,024, MAE 0,018, MAPE 0,400 olmuştur. Bu çalışma sonucunda ÇKA’nın başarısının geleneksel ekonometrik bir yöntem olan doğrusal regresyona göre daha iyi olduğu sonucuna ulaşılmıştır. Elde edilen bu sonuçlar Türkiye’de bundan önceki yapılan çalışmalar ile paralellik göstermektedir. Ayrıca bu çalışma profesyonel fon yöneticiler için de yönettikleri fonların performansını sürdürülebilir hale getirmesi ve değişen ekonomik koşullara göre portföylerini oluşturmasında yardımcı olacağı düşünülmektedir.