Sürdürülebilir Kalkınma Amaçları
Yayınlar
- MakaleAçık Erişim2025Scopus SJRQ1Hybrid GA-ConvLSTM for data-driven prediction of climate variables: a case study of the most biodiverse cities in India
Utku, Anıl; Akyol, Sinem
Climate variables affect ecosystems’ structure, sustainability, and biodiversity. Monitoring and analysis of climate variables are essential for the management of ecosystems and the conservation of biodiversity. Temperature affects the distributions, habitats, and phonological events of living things in ecosystems. Humidity is essential for meeting plants’ water needs and the habitat’s quality and health. Dew point is essential for water quality, balance, plant health, and microclimate. Wind speed is essential for pollinating plants, microhabitats, and the health of living things. For such reasons, accurate prediction of climate variables using artificial intelligence methods can effectively predict the future of ecosystems and biodiversity and develop future strategic plans. This study aimed to predict the climate variables of temperature, humidity, dew point, and wind speed of Bhubaneshwar, Dehradun, Guwahati, Mysore, and Nagpur, among the cities with the highest biodiversity in India. It aimed to determine the long-term impacts of climate change on managing ecosystems and biodiversity protection. The hyperparameters of the ConvLSTM model, developed using CNN and LSTM, were optimized with a genetic algorithm, and the GA-ConvLSTM hybrid model was created. GA-ConvLSTM was tested with XGBoost, SVM, ANFIS, CNN, LSTM, and base ConvLSTM using approximately 15 years of hourly, up-to-date, and real-time climate variable data of cities. This work is one of the pioneering studies using GA to optimize deep learning hybrid models for multi-city climate prediction through extensive long-term high-resolution datasets. The experimental results reveal that GA-ConvLSTM outperforms conventional models, producing an R2 score exceeding 0.9 for each city across all climate variables. GA-ConvLSTM produced temperature predictions with 0.833 RMSE and 0.995 R2 for Bhubaneswar, humidity predictions with 3.461 RMSE and 0.990 R2, dew point predictions with 0.632 RMSE and 0.995 R2, and wind speed predictions with 2.278 RMSE and 0.993 R2. The experimental findings confirm the improved performance over conventional models due to SVM and XGBoost failing to achieve these R2 values and high RMSE scores. The research meets three principal obstacles because of the combination of stochastic climate datasets and complex nonlinear patterns, which also encounter limitations in conventional methods for managing scale-dependency relationships and manual parameter adjustments. This approach delivers three main strengths because it combines CNN and LSTM networks through genetic programming while tracking temporal and spatial relationships and processing noisy data automatically, leading to improved results compared to other methods. Unlike other published models in the field, the hybrid model GA-ConvLSTM combines optimized CNN and LSTM elements with genetic algorithm elements. The proposed model received its first application in this research for multi-city climate prediction through training with 15 years of hourly data. Standard methods encounter challenges when detecting long-term dependencies, but GA-ConvLSTM successfully addresses this issue.
AtıfScopus1OpenAlex1
İncele - MakaleAçık Erişim2024Hybrid ConvLSTM Model for Wind Speed Prediction: A Case Study for India's Highest Wind Speed Cities
Utku, Anıl; Akyol, Sinem
Wind speed prediction is very important for logistics, energy production and renewable energy sources, aviation and maritime, agriculture, disaster management, environmental monitoring, construction, life planning and economic activities. In this study, it was aimed to predict the wind speeds of Jaisalmer, Kochi, Mangalore, Puri and Rameswaram, which are among the cities with the highest wind speeds in India. ConvLSTM hybrid model was developed by taking advantage of the effective features of CNN and LSTM models for wind speed prediction. With ConvLSTM, it is aimed to better determine the dynamics of wind speed by processing spatial and temporal data simultaneously. ConvLSTM has been extensively tested using an approximately 15-year hourly dataset consisting of RF, SVM, ANFIS, CNN and LSTM and wind speed measurements at 10 meters and 100 meters’ altitudes. Experimental results show that ConvLSTM has R2 values above 0.9 for almost all of the city and wind speed parameters, and outperforms the compared models with the lowest error values in terms of RMSE and MAE error metrics.
AtıfOpenAlex1TR Dizin1
İncele - MakaleAçık Erişim2020Scopus SJRQ2Sentiment classification within online social media using whale optimization algorithm and social impact theory based optimization
Akyol, Sinem; Alatas, Bilal
In recent years, social networks have become an important part of daily life and are becoming increasingly important. In this study, sentiment analysis or opinion mining, which is one of the most well-known social network analysis problems, is considered as an optimization problem for the first time. In the same way, for the first time, sentiment analysis is considered as a multi-objective problem. Whale Optimization Algorithm and Social Impact Theory based Optimization Algorithm, which are the current intelligent optimization algorithms, are adapted for the sentiment analysis problem. Furthermore, memory feature is integrated into Social Impact Theory based Optimization Algorithm in order to obtain effective results in this study. The obtained results from the proposed algorithms are also compared with thirty-three supervised learning algorithms within real IMDB, Polarity, and Amazon data sets. In order to evaluate the performance of the results; accuracy percentage, precision, recall, F-Measure, and MCC that are the five most commonly used measure in the literature are used. When the results are examined, it is observed that adapted metaheuristic optimization algorithms give more successful results in sentiment analysis problem.
AtıfScopus35WoS27OpenAlex43
İncele - MakaleAçık Erişim2017Scopus SJRQ1Plant intelligence based metaheuristic optimization algorithms
Akyol, Sinem; Alatas, Bilal
Classical optimization algorithms are insufficient in large scale combinatorial problems and in nonlinear problems. Hence, metaheuristic optimization algorithms have been proposed. General purpose metaheuristic methods are evaluated in nine different groups: biology-based, physics-based, social-based, music-based, chemical-based, sport-based, mathematics-based, swarm-based, and hybrid methods which are combinations of these. Studies on plants in recent years have showed that plants exhibit intelligent behaviors. Accordingly, it is thought that plants have nervous system. In this work, all of the algorithms and applications about plant intelligence have been firstly collected and searched. Information is given about plant intelligence algorithms such as Flower Pollination Algorithm, Invasive Weed Optimization, Paddy Field Algorithm, Root Mass Optimization Algorithm, Artificial Plant Optimization Algorithm, Sapling Growing up Algorithm, Photosynthetic Algorithm, Plant Growth Optimization, Root Growth Algorithm, Strawberry Algorithm as Plant Propagation Algorithm, Runner Root Algorithm, Path Planning Algorithm, and Rooted Tree Optimization.
AtıfWoS188OpenAlex218İncele - MakaleYalnız künye2016Scopus SJRQ2Kedi sürüsü optimizasyon algoritmasıyla doğru ve anlaşılabilir nümerik sınıflandırma kurallarının otomatik keşfi
Alatas, Bilal; Akyol, Sinem
Metasezgisel algoritmalar, büyük boyutlu optimizasyon problemleri için, kabul edilebilir sürede optimuma yakın çözümler verebilen algoritmalardır. Genel amaçlı metasezgisel optimizasyon algoritmaları; biyoloji tabanlı, fizik tabanlı, sürü tabanlı, sosyal tabanlı, müzik tabanlı, spor tabanlı ve kimya tabanlı olmak üzere yedi farklı grupta değerlendirilmektedir. Sürü zekâsı tabanlı optimizasyon algoritmaları kuş, balık, kedi ve arı gibi canlı sürülerinin hareketlerinin incelenmesiyle geliştirilmiştir. Veri madenciliği, büyük ölçekli verilerden anlamlı ve faydalı bilginin keşfedilmesi işlemidir. Sınıflandırma kurallarının madenciliği üzerinde en çok çalışılan veri madenciliği problemlerinden biridir ve bu problemde veri kümelerinden kullanıcıların rahatça anlayabileceği kurallar çıkarılmaktadır. Bu çalışmada, en güncel sürü zekâsı optimizasyon algoritmalarından Kedi Sürüsü Optimizasyonu (KSO) bildiğimiz kadarıyla ilk kez nümerik ya da karışık tipte verilerden oluşan veritabanlarında sınıflandırma kurallarının otomatik keşfinde kullanılmıştır. Kuralda yer alabilecek ilgili niteliklerin doğru aralıkları için bir ön işlem kullanılmamış, kurallarla birlikte otomatik olarak bulunması da KSO tarafından sağlanmıştır. Ayrıca, kullanılan amaç fonksiyonu çok esnektir ve farklı amaçlar kolaylıkla fonksiyona entegre edilebilir. Bu amaçla, UCI veri ambarından alınan 4 adet veritabanı kullanılmış; doğru ve anlaşılabilir sınıflandırma kuralları keşfedilmiştir. Bulunan sonuçlarla Weka programından elde edilen sonuçlar karşılaştırılmıştır. KSO algoritması üzerinde herhangi bir iyileştirme yapılmadığı ve bu alanda ilk kez kullanıldığı halde, elde edilen sonuçların umut verici olduğu gözlenmiştir.
AtıfTR Dizin7İncele - MakaleAçık Erişim2016Scopus SJRQ2Automatic mining of accurate and comprehensible numerical classification rules with cat swarm optimization algorithm
Akyol, Sinem; Alatas, Bilal
Optimization is the process of finding the best solution of a problem. There are many optimization algorithms proposed for optimization problems. The metaheuristic algorithm can give solutions close to optimum in an acceptable period of time for large-scaled optimization problems. The metaheuristic optimization algorithms are evaluated in seven different groups which are biology-based, physics-based, swarm-based, social-based, music-based, sport-based, and chemistry-based. The swarm-based optimization algorithms have been developed by observing the behaviors of creatures e.g. birds, fishes, cats, bees etc. Data mining is the process of discovery of meaningful and useful data within huge databases. Classification rules mining is one of the most commonly studied data mining problems and with methods for this problem, users can easily understand the rules extracted from databases. In this work, one of the most recent swarm based optimization algorithm, Cat Swarm Optimization (CSO), has been firstly used for classification rules mining within databases composed of numerical or mixed data types automatically. There is not any preprocess for finding true ranges for appropriate attributes of the rules, this has been automatically done by CSO. Furthermore, the used objective function is very flexible and many different objectives can easily be added to. For this purpose, four numerical databases obtained from UCI data warehouse have been used and accurate and comprehensible classification rules have been mined. The results have been compared with the results obtained from Weka program. Although CSO has not been embedded with any improvement and has firstly implemented in this research area, the obtained results seem promising.
AtıfWoS9OpenAlex4İncele