Over the course of the past three years, I have been actively engaged in the fields of data science and artificial intelligence. I have spent half of this duration developing machine learning solutions for the finance and tourism sectors, and the other half focused on the game industry. During this time, I also successfully completed a master degree, which has further equipped me with a solid understanding of the underlying principles and techniques of these fields. My expertise is grounded in statistics and machine learning, and encompasses a range of skills including working with large and complex datasets, designing and executing AB tests, building predictive models, and conducting statistical analysis. My experience in these areas has been honed through my involvement in various projects across different domains. As of April 2023, I have taken up the role of Senior Data Scientist at Zynga, where I am applying my knowledge and skills to develop and deliver innovative solutions in the realm of data science and artificial intelligence.
Contribute to the design, monitoring, analysis, and clear communication of A/B test results. Take proactive measures to develop novel analyses and tools utilizing statistical and machine learning models, enabling data-driven decision-making processes and the formulation of effective strategies for improving retention, monetization, and ROI. Offer valuable business analytical insights during discussions and actively participate in driving product decisions and prioritization within the mobile gaming sector, encompassing popular titles like Merge Dragons!, Merge Magic!, and Pirate Evolution!
I have experience in developing cutting-edge artificial intelligence and natural language processing solutions for the purpose of evaluating user experience in the tourism industry. Specifically, my work includes the creation of a distributed, bidirectional, cross-reputation, and aspect-based review service for hotels and tourism-related entities. This solution can be utilized across multiple countries. My work in this field includes conducting morphological and dependency analysis, sequence tagging, parser development, aspect-based categorization, and cross-reputation analysis
SummaryWe introduced novel document recommendation system and visual exploratory tool that utilizes a combination of natural language processing algorithms and graph representation learning methods. The system leverages conjoined n-grams, temporal information, and additional meta-data to create a heterogeneous graph network representation, effectively solving the cold-start problem
An explorative visual analytics tool which enhanced by NLP methods is designed to explore and learn from texts.
Time series were transferred to the frequency domain by using Wavelet Decomposition. System is designed by using spectrograms that created from frequency-domain features are given as inputs to deep learning algorithms, such as CNNs and LSTMs.
Deep-Learning based architecture for hierarchical text classification designed and developed for the SemEval-2020 competition on Multilingual Offensive Language Identification in Social Media ( OffensEval 20) where we achieved 2nd position in Turkish sub-task and 6th in Greek . Manuscript accepted and published in September 2020.
System is designed for identifying the semantic overlap between sentences.Paraphrase identification with machine learning and deep learning.
BeautyGan is designed for the solution to the facial make-up translation between images using Generative adversarial network .
Perform comprehensive analysis and forecasting on time-stamped data.Accomplish successful predictions over automobile sales volume in Turkey. Manuscript accepted and published by IDSES 2019 - International Data Science and Engineering Symposium.
You can reach me via email at anilozdemir9696@gmail.com.