Under the influence of the COVID-19 pandemic and the concurrent oil conflict between Russia and Saudi Arabia, oil prices have exhibited unusual and sudden changes. For this reason, the volatilities of the West Texas Intermediate (WTI), Brent and Dubai crude daily oil price data between 29 May 2006 and 31 March 2020 are analysed. Firstly, the presence of chaotic and nonlinear behaviour in the oil prices during the pandemic and the concurrent conflict is investigated by using the Shanon Entropy and Lyapunov exponent tests. The tests show that the oil prices exhibit chaotic behavior. Additionally, the current paper proposes a new hybrid modelling technique derived from the LSTARGARCH (Logistic Smooth Transition Autoregressive Generalised Autoregressive Conditional Heteroskedasticity) model and LSTM (long-short term memory) method to analyse the volatility of oil prices. In the proposed LSTARGARCHLSTM method, GARCH modelling is applied to the crude oil prices in two regimes, where regime transitions are governed with an LSTAR-type smooth transition in both the conditional mean and the conditional variance. Separating the data into two regimes allows the efficient LSTM forecaster to adapt to and exploit the different statistical characteristics and ARCH and GARCH effects in each of the two regimes and yield better prediction performance over the case of its application to all the data. A comparison of our proposed method with the GARCH and LSTARGARCH methods for crude oil price data reveals that our proposed method achieves improved forecasting performance over the others in terms of RMSE (Root Mean Square Error) and MAE (Mean Absolute Error) in the face of the chaotic structure of oil prices.

A b str a c t: T his study presents a novel hybrid Turkish text sum m arization system that com bines structural and semantic features. T h e system uses 5 structural features, 1 o f w hich is newly proposed and 3 are semantic features whose values are extracted from Turkish W ikipedia links. T h e features are com bined using the weights calculated by 2 novel approaches.

No abstract

In this paper, we propose hybrid models for modelling the daily oil price during the period from 2 January 1986 to 5 April 2021. The models on S2 manifolds that we consider, including the reference ones, employ matrix representations rather than differential operator representations of Lie algebras. Firstly, the performance of LieNLS model is examined in comparison to the Lie-OLS model. Then, both of these reference models are improved by integrating them with a recurrent neural network model used in deep learning. Thirdly, the forecasting performance of these two proposed hybrid models on the S2 manifold, namely Lie-LSTMOLS and Lie-LSTMNLS, are compared with those of the reference LieOLS and LieNLS models. The in-sample and out-of-sample results show that our proposed methods can achieve improved performance over LieOLS and LieNLS models in terms of RMSE and MAE metrics and hence can be more reliably used to assess volatility of time-series data.

In this work, a face recognition system working on video records is constructed. The aim of this study is to analyze; how -instead of using all video frames and building a complex modelusing only a subset of informative frames (named representative frames), automatically selected and preprocessed by a system, affects success rate and processing time.The basic idea in selecting the representative frames is that; faces that are captured from the front contain sufficient information for recognition. So a simple algorithm -that uses facial features (eyes and mouth) and positions of these features on the face-can be developed to extract these frames from the videos. Even the number of these frames is relatively small, handling these frames in a proper way and preprocessing them with some simple image processing techniques effect the recognition rate positively.After representative frames are extracted and preprocessed, dimensional analyses are applied and extracted data is transformed onto a new space.Finally, transformed data is used to build and use a trainer, by which the system is trained with videos consisting of only one person per each, however having no restriction in pose, angle and rotation. Then different videos with same situations are presented to the trained classifier for recognition.

No abstract

Java has been a widely used programming language for a long time in various fields. Java and its libraries have been frequently updated for various reasons including bugs, change requests, performance and usability requirements and so on. In this paper, we examine how these changes affect the use of Java and analyze trends in its usage. As a data source, we use the Stack Overflow public dataset which is the largest online Q&A site about software technologies. We firstly employ a practical approach to extract the Javarelated posts from the Stack Overflow dataset using cosine similarity and compare it with previous works. We then apply Latent Dirichlet Allocation (LDA) to the corpus for topic modelling. We divided the data set into two-year periods to obtain consistent clusters. After obtaining main topics, we examine topics and keywords on a two-year basis. Finally, unlikely previous works, we manually classify topics into two as "domain-specific" and "development environment" and investigate tendencies of these classes to change in both the short term and the long term.

Özetçe-Çalışmada, Türkiye'de faaliyet gösteren bir banka için müşteri terk modeli geliştirilmiştir. Bankacılık sektöründe müşteri banka ilişkisinin süresi bir kontrata dayalı olmadığından, bu modeli bankacılık için geliştirmek diğer sektörlere oranla daha zorlu bir süreçtir. Model oluşturmak için öncelikle müşteri ham verileri kullanışlı ve anlamlı bir hale dönüştürülerek veri ambarına alınması çalışması yapılmıştır. Daha sonra hazırlanan bu veri kümesi üzerinde, veri madenciliği teknikleri kullanılarak bir terk tahmin modeli geliştirilmiştir. Geliştirilen modellerin tahminleme performansları doğruluk oranı, duyarlılık, belirginlik, kappa istatistiği ve AUROC ölçümleri kullanılarak değerlendirilmiştir. Anahtar Kelimeler -Müşteri terk tahmini, müşteriyi elde tutma, CRM, veri madenciliği, karar ağaçları, yapay sinir ağları, lojistik regresyon, rastgele ormanAbstract-This paper proposes a customer churn model for a private bank in Turkey. It is more challenging to put forth a model for banking sector as there are no contractual agreements between a customer and a bank regarding the duration of services. During the development of the model, we first converted the raw data into a usable and meaningful form. Later, using data mining techniques on our data, we have developed a "churn prediction model". Prediction performance is evaluated using accuracy, sensitivity, specificity, kappa statistic and area under the curve based measures.