AMAhmet Anil Mungen

Ahmet Anil Mungen

*This is an algorithmically generated author record based on publicly available information.

Ahmet Anil Mungen has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Regression Analysis of Age of Empires II DE Match Results with Machine Learning

    by Iclal Cetin Tas, Ahmet Anıl Mungen

    This study analyzes the effect of the nationalities chosen by the players and the maps played on the match results in the popular e-sport game Age of Empires II. In the study, a prediction model was created using machine learning algorithms on a large dataset. The model can predict the probability of winning if a player or a team chooses a particular nation on a particular map. This analysis covers both individual and team play and offers players and e-sport players a guide to consciously guide their strategic decisions. According to the results obtained, it was seen that the best results were obtained by using the GPR algorithm with RMSE value of 0.50.

  • Open Access

    Sunucuların Anomali Durumlarinin Yapay Zeka Metotları ile Tahmin Edilmesi

    by Mehmet Fatih Savran, Ahmet Anıl MÜNGEN

    Sunucularda anormallik tespiti belirli bir metot ve uygulama yoluyla aykırı değerlerin diğerlerinden ayrıştırılarak analiz edilmesiyle oluşan ve genelde olağan dışı durumları tespit etmekte kullanılır. Anormalliklerin erkenden tespit edilmesi ön görülebilirlik kararlar vermeyi ve gerekirse savunma mekanizması geliştirilmesinde kullanılabilir. Önemli bir problem olarak bilinen Anormallik Tespiti birçok tarama ve uygulama sahasında araştırılmaktadır. Genelde araştırmacılar bu bahsi geçen probleme yapay zeka, makine öğrenimi ve durum makine modellemesi gibi teknikleri kullanarak çözüm arayışına girmişlerdir. Sunucuların anormallik testleri ve analizi yapılabilir ve bu yöntem-teknikler kullanılarak çıkarımlar yapılabilir. Sunuculardan alınan CPU, Network, Disk, Memory değerleri anomali testinde kullanılmak üzere veri analiz aşamalarından geçer ve teknikler uygulanarak modellemesi yapılır. Bu çalışmada toplanan veri kümesi kullanılarak YSA, Karar Ağacı, Rastgele Orman, K- En Yakın Komşu ve Ekstra Karar Ağacı algoritmalarının anormali tespit performansları test edilmiştir. Yapılan testlerde anormal durumlarının belirlenmesinde % 99.94 oranıyla YSA’nın başarılı olduğu görülmüştür. Önerilen yöntem, toplanan veri ve önerilen yöntemin diğer yöntemler ile karşılaştırmalı analizleri çalışma içerisinde sunulmuştur.

  • Personalised publication recommendation service for open-access digital archives

    by Ahmet Aníl Müngen

    Increase in the number of open-access academic publications and open-access institutional academic archives led more researchers use these archives. No model offers personalised publication suggestions in academic archives. A central service architecture has been proposed towards personalised academic article recommendations for open-access digital archives. Thus, it has been possible to make personalised suggestions for open-access digital archives and enable researchers to discover new publications. A service based on the centralised micro-service architecture model was proposed in the study. Also, TF-IDF and article classification methods were used together for the personalised publication suggestion system. For the first time globally, the proposed method was used with 1464 real users in 49 open-access archives. F-measure success value was found to be higher than 0.8 for recommended publications.


  • A General Review of the Methods Used Financial Time Series Forecasting

    by Nuh Yurduseven, Ahmet Anıl MÜNGEN

    Geçmişte olduğu gibi günümüzde de yatırımcılar için finansal verilerin trendinin tahmin edilebilmesi ve bu bilgi kullanılarak bir finansal strateji oluşturulması oldukça önemlidir. Fakat günümüzde hızlı internet bağlantıları ile finansal verilerin hızlı ulaşması ve bilişim ve bulut sistemlerindeki gelişmeler, finansal tahminlemek için yapay zekâ algoritmalarının kullanılması bu alanda rekabeti artırmaktadır. Fintech içinde portföy yönetimi gibi alanlarda yapay zekâ uygulamalarının kullanım payı gittikçe artmaktadır. Bu çalışmanın amacı finansal zaman serisi verileri tahminlemek için yapılan daha önceki akademik çalışmaları derlemek, zaman serilerinin tahmin etmek için kullanılan yapay zekâ algoritmalarını açıklamak ve tahmin edilen bazı finansal veri tiplerini ve bağımlılıklarını irdelemektir. Çalışma sonunda incelenen makalelerde kullanılan tekniklerin yeterlilikleri ve hangi veri tipi için hangi metodun daha başarılı sonuçlar verebileceği gibi çıkarımlar yapılmıştır.


  • Open Access

    YAPAY SİNİR AĞLARI ve DESTEK VEKTÖR MAKİNELERİ YÖNTEMLERİ ile BÖLGESEL TRAFİK YOĞUNLUK TAHMİNİ

    by İclal Çeti̇n Taş, Ahmet Anıl MÜNGEN

    Büyük şehirlerde kilometre başına düşen insan yoğunluğu arttıkça trafik sıkışıklığı artmakta ve yolcuların daha fazla sürelerini trafikte harcamaktadırlar. Trafik sıkışıklığı nedeni ile harcanan ekstra zaman ve yakıt hem kullanıcılar hem de ülkeler için büyük bir gider kalemidir. Büyükşehirlerde yaşayan vatandaşların trafik yoğunluğunun zaman bazlı değişimini tahmin etmek ve buna göre planlama yapmak bir zorunluluk haline dönüşmüştür. Trafik sıkışıklıkları genelde tüm şehirde aynı anda gerçekleşmez. Bölgesel olarak yaşanan trafik sıkışıklıkları diğer yolları da etkilemesi ile yaygınlaşır. Bu çalışma da yapay sinir ağları (YSA) kullanılarak önerilen yöntem ile geçmiş trafik verileri kullanarak bölgesel yoğunluklar tahmin edilmeye çalışılacaktır. Çalışma birçok muadil çalışmanın dışında hava durumu gibi çevresel etkenleri de alarak tahmin modellemesinin başarısını arttırmaktadır. İstanbul Büyük Şehir Belediyesi Açık Veri Portalından toplanan 75 farklı noktaya ait 150.000 veri kullanarak önerilen model test edilmiş ve yaklaşık %90 başarı ile bölgesel trafiğin yoğunluğunu tespit edilebilmiştir.


  • Finding the Relationship Between News and Social Media Users’ Emotions in the COVID-19 Process

    by Ahmet Anıl MÜNGEN, İ̇̇rfan Aygün, Mehmet Kaya

    Nowadays, social media and online sharing sites are frequently used to share thoughts about daily events. Thanks to the posts made by internet users on these platforms, first, quite big data is generated to interpret the agenda. More than 10,000 comments of more than 5000 users made about COVID-19 from online websites between 15 March and 15 May were collected in this study. Then, emotional analysis on these comments was carried out with BERT, GRU, LSTM and TF-IDF methods. The changes in the amount of user comments and the emotions reflected by the comments have been associated with the actual events of these dates. It has been determined which types of events affect users more. In addition, the emotional response changes of the users to the official COVID-19 statistics were measured and the peak points of the emotional changes were determined. Finally, the emotion classification methods applied were evaluated by user questionnaires and their successes were determined according to F-Measure.


  • Friend Recommendation Decision Systems via Multiple Social Network Alignment

    by Ahmet Anil Mungen, Betul Bulut, Mehmet Kaya

    Today, almost all internet users have more than one social network account on different social networks for interaction with friends and other users. Gathering data from various networks to combined into a single node can be used for increasing the success rate of recommendation systems. In this study, data related to thousands of users in nine different social networks are used for successful recommendations to the users. The anchor method is used for topological alignment, and the relationship between nodes is taken into account for calculation. Also, the node similarity method is used to increase the success rate. In this method, the number of successful node matching is increased thanks to the feature selection criteria. An original node alignment and node similarity methods are proposed in the study. Because of combine both node alignment and node similarity method, the proposed method is very successful for the friend recommendation.


  • Exploring the Relationship Between Academicians via Reference Parsing

    by Gizem Cay, Ahmet Anil Mungen, Emre Dogan, Mehmet Kaya

    Although the citation number is significant for a paper, it is important to determine the value of citation with the information of the citations' owner and their value for revealing the value of the citations. Our study proposes a method for parsing the reference sections of an academic publications. It will provide a graphical representation of the relationships of the people who give and cite the resulting shredding works in a weighted directional graph. Sobiad Citation Index and TUBİTAK Dergipark and ULAKBIM TRDizin were used as the data set. The proposed method was found to be successful and result show in the graphs and experimental results.

  • Features and Connection Based Network Alignment Method

    by Ahmet Anil Mungen, Mehmet Kaya

    In almost all proposed computer science algorithms, success rates are directly related to the data set. Recommendation and link prediction methods are the most critical methods in which the success of the repository directly affects the success of the method. The cold-start problem can be classified as basically a missing data problem. Since there are dozens of social networks in the world, not being able to collect all data in one source can be called as missing data problem. The fact that the users' Twitter pages cannot be used as a single source without other social networks like Instagram for collecting data about users. Instead of this, it can be more successful if data from both networks to connect. It is not usual for users to link accounts across different social networks or merge them into a single account. This study aimed to find the same users in different social networks by looking at the attributes of the profile, their connections, and shares on the network. In this study, based on the status and shares of the users on the network, the proposed method has been successfully found in the profiles of the person in different networks.


  • Generation of Original Text with Text Mining and Deep Learning Methods for Turkish and Other Languages

    by Emre Dogan, Buket Kaya, Ahmet Mungen

    The amount of content on the web has increased dramatically since the Internet began providing users with the ability to produce content. Initial work on original text production has aimed at publishing the given data by putting in a certain mold. The most obvious example of this is the analysis reports on sporting events. However, preparing an original text compiled with general information about a subject has become a subject of interest to scientists as well. Although Neural Networks and Markov models were used previously for original text production, the original text generation process and comparison of the success rates weren't done using the Turkish language and the academic publication data repository dataset. In this study, it was tried to create summary information/original content about a specific subject by using Wikipedia TR for the Turkish language and the data pool created with hundreds of thousands of academic publications. In the study, texts were produced with Markov Model and LSTM, which were previously proposed, and the results are comparatively shared in detail. In the evaluation study, the performance of the proposed method was examined, and the correctness of the techniques was evaluated concerning syntactic accuracy and semantic preservation. The results are evaluated by presenting a mixture of original and machine-generated texts to the actual user for the success test of the proposed method. The success rate of the results is calculated with accuracy, recall, and f-measure. The results are very promising because it has been observed that the method can produce accurate and quality representations.