Majör koin verileri kullanılarak alt-token'ların fiyatlarının makine öğrenimi modelleri ile tahmini
Date
2024-07-26
Authors
Apak, Muhammet
Journal Title
Journal ISSN
Volume Title
Publisher
Bursa Uludağ Üniversitesi
Abstract
Bu tez çalışmasının amacı, majör kripto paraların (örneğin Ethereum, Solana ve Binance Koin) kendi ekosistemlerinde inşa edilen altkoinlerin fiyatları üzerindeki etkilerini analiz ederek bu ekosistemlerin dinamik yapısını ve piyasalar üzerindeki etkilerini incelemektir. Kripto ekosistemlerinde majör koinlerin altkoinlerin fiyatlarına olan etkilerini belirlemek için LSTM-CNN hibrit modeli kullanılmıştır. Bu çalışmada kullanılan veri setleri Ethereum, Solana ve Binance Koin tokenlarına ait fiyat verilerini içermektedir. Veri setleri, CoinMarketCap ve Binance gibi güvenilir kaynaklardan elde edilmiştir. LSTM-CNN hibrit modeli kullanılarak ilgili majör koinlerin ekosistemlerindeki altkoinlerin fiyatları tahmin edilmiştir. Model eğitimi ve testi için veri ön işleme, özellik mühendisliği ve modelin performansını değerlendiren hata metrikleri kullanılmıştır. Modelin performansı, RMSE ve MAPE gibi metriklerle değerlendirilmiştir. Tez çalışmasının bulgularına göre, majör koinlerin fiyat hareketleri altkoinlerin fiyatları üzerinde belirgin etkiler yaratmaktadır. Özellikle Ethereum, Solana ve Binance Koin ekosistemlerinde yapılan analizlerde, majör koinlerin fiyat değişimlerinin altkoinlerin fiyatlarını önemli ölçüde etkilediği gözlemlenmiştir. LSTM-CNN hibrit modeli, bu ilişkileri başarılı bir şekilde tespit etmiş ve tahmin performansı yüksek sonuçlar vermiştir. Modelin performans metrikleri, tahminlerin doğruluğunu ve modelin güvenilirliğini göstermektedir. Bu tez çalışması, majör kripto paraların kendi ekosistemlerinde inşa edilen altkoinlerin fiyatları üzerindeki etkilerini başarılı bir şekilde analiz ederek kripto ekosistemlerinin dinamik yapısını ve piyasa üzerindeki etkilerini anlamayı başarmıştır. Elde edilen sonuçlar, kripto para piyasalarının daha iyi anlaşılmasına ve yatırım stratejilerinin geliştirilmesine önemli katkılar sağlamaktadır. Bu bağlamda, çalışma kripto para piyasalarının dinamiklerini anlamak ve bu alandaki literatüre katkıda bulunmak açısından değerlidir.
The purpose of this thesis is to analyze the impact of major cryptocurrencies (such as Ethereum, Solana, and Binance Coin) on the prices of sub-tokens built within their ecosystems, thereby examining the dynamic structure of these ecosystems and their effects on the markets. A hybrid LSTM-CNN model has been employed to determine the effects of major coins on the prices of sub-tokens within their respective ecosystems. The datasets used in this study contain price data related to Ethereum, Solana, and Binance Coin tokens. These datasets have been obtained from reliable sources such as CoinMarketCap and Binance. The prices of sub-tokens in the ecosystems of the related major coins have been predicted using the hybrid LSTM-CNN model. Data preprocessing, feature engineering, and error metrics evaluating the model's performance were utilized for model training and testing. The performance of the model has been assessed using metrics such as RMSE and MAPE. According to the findings of the thesis, the price movements of major coins have significant effects on the prices of sub-tokens. Particularly in the ecosystems of Ethereum, Solana, and Binance Coin, analyses have shown that price changes of major coins significantly impact the prices of sub-tokens. The hybrid LSTM-CNN model has successfully identified these relationships and yielded high prediction performance. The performance metrics of the model demonstrate the accuracy of the predictions and the reliability of the model. This thesis has successfully analyzed the impact of major cryptocurrencies on the prices of sub-tokens built within their ecosystems, thereby understanding the dynamic nature of cryptocurrency ecosystems and their effects on the market. The results contribute significantly to a better understanding of cryptocurrency markets and the development of investment strategies. In this context, the study is valuable for understanding the dynamics of cryptocurrency markets and contributing to the literature in this field.
The purpose of this thesis is to analyze the impact of major cryptocurrencies (such as Ethereum, Solana, and Binance Coin) on the prices of sub-tokens built within their ecosystems, thereby examining the dynamic structure of these ecosystems and their effects on the markets. A hybrid LSTM-CNN model has been employed to determine the effects of major coins on the prices of sub-tokens within their respective ecosystems. The datasets used in this study contain price data related to Ethereum, Solana, and Binance Coin tokens. These datasets have been obtained from reliable sources such as CoinMarketCap and Binance. The prices of sub-tokens in the ecosystems of the related major coins have been predicted using the hybrid LSTM-CNN model. Data preprocessing, feature engineering, and error metrics evaluating the model's performance were utilized for model training and testing. The performance of the model has been assessed using metrics such as RMSE and MAPE. According to the findings of the thesis, the price movements of major coins have significant effects on the prices of sub-tokens. Particularly in the ecosystems of Ethereum, Solana, and Binance Coin, analyses have shown that price changes of major coins significantly impact the prices of sub-tokens. The hybrid LSTM-CNN model has successfully identified these relationships and yielded high prediction performance. The performance metrics of the model demonstrate the accuracy of the predictions and the reliability of the model. This thesis has successfully analyzed the impact of major cryptocurrencies on the prices of sub-tokens built within their ecosystems, thereby understanding the dynamic nature of cryptocurrency ecosystems and their effects on the market. The results contribute significantly to a better understanding of cryptocurrency markets and the development of investment strategies. In this context, the study is valuable for understanding the dynamics of cryptocurrency markets and contributing to the literature in this field.
Description
Keywords
Altkoin, Makine öğrenimi, LSTM, CNN, Kripto para, Fiyat tahmini, Kripto para ekosistem, Ethereum, Solana, Binance Koin, Majör koin, Major coin, Sub-token, Machine learning, Cryptocurrency, Price prediction, Cryptocurrency ecosystem, Binance Coin.