adjlsv@r.postjobfree.com ibrahimsevkieldivan ibrahimsevki Lodi, NJ 07644 201-***-****
IBRAHIM S ELDIVAN,
SUMMARY
Have 10+ years of experience in data-driven fields with a strong background in statistics, fast learning data scientist, natural language processing, big data, and machine learning. I am passionate about solving real-life problems using statistical methods, building predictive models, and getting valuable insight from the data. TECHNICAL SKILLS
Tools - Python, SQL, Alteryx, Jupyter Notebook, NumPy, Pandas, SciPy, Sci-kitLearn, Pyspark, Statsmodel, TensorFlow, NLTK, Seaborn, Matplotlib, GitHub/Git, Spark, AWS Cloud Services, Pytorch, Tensorflow, Tableau Machine Learning - Regressions, Classification, Clustering, Hypothesis Testing, Supervised/Unsupervised, Bagging (Random Forest), Boosting
(AdaBoost, Gradient Boosting, XGBoost, CatBoost) algorithms, Deep Learning, Transfer Learning, Visualization - Seaborn, Matplotlib, Plotly, Tableau Big Data - Shell Scripting, Docker & Kubernetes, PostgreSQL, Hadoop HDFS/YARN, Hive, Sqoop, Data Generator library, Kafka, Spark, CI/CD Jenkins, Model Deployment
Statistics - Hypothesis Testing, Principal Component Analysis (PCA), Cross-Validation, Regularization, ANOVA, N-Vivo, A/B Testing PUBLICATIONS
• Pathway From Tranquility to Violent Radicalization: A Case Study of 2003 Istanbul Bombings 2011
• Analyzing Terrorism Using Special Analysis Techniques: A Case Study of Turkish Cities 2011 Ph.D.
No Visa Sponsorship Required Fully Work Authorized Open to Relocation PROJECTS
• Sentiment Analysis of Top Trending Subjects on Twitter in The US 2020
• Used Car Price Allocation (Regression) 2020
• COVID-19 Patient Recovery Prediction Model Using Random Forest Classifier 2020
• Adult Income Prediction with Classification Algorithms 2020
• Discriminant Features of Violent and Non-Violent Al-Qaeda Member (Discriminant Function Analysis/Classification Model 2011
• Decision Tree Specialty Coffee Case Study (Classification) 2020 EDUCATION
Rutgers University - Newark, New Jersey
Ph.D. in Criminal Justice
John Jay College - New York City, New York
Master of Science in Criminal Justice
Institute of Security Sciences - Ankara, Turkey
Bachelor's in Security Sciences
May 2011
Aug 2008
Jun 1997
WORK EXPERIENCE
• Taught Data Analysis & Statistics and Research Methods to graduate and undergraduate students
• Took part in a grant project and granted $716K from the NIJ for Domestic Terrorism in the US with Risk Terrain Modeling
• Performed Time Series Analysis, data compilation, data analysis with SPSS, and data interpretation
• Analyzed multiple intelligence analysis, for experimental design, statistical robustness, using PCA, Regression, Random Forest-based model.
• Conducted Stepwise Discriminant Function Analysis to predict violent and nonviolent activities
• Redesigned data-oriented intelligence forecasting process and conducted data analysis training courses to field-officers,
• Increased accurate-intelligence rates by 30%, reduced training costs per field-officer by 12%, get 85% positive feedbacks from trainees about redesigned data-oriented intelligence forecasting process Data Scientist (Contract)
• Applied Gradient Boosting tree-based algorithm on real-time crime data to predict the occurrence of the non-violent incident within the given timeframe with 0.83 F1 score which is 9% better than the pre-existing solution based on previous studies
• Created time-based supervised ML model for crime prediction, combined with unsupervised clustering hotspots and ARIMA models to predict city-wise point estimations with 87% precision
Peel9 Records Management & Analytics, Cincinnati, USA (remote) May 2020 - Oct 2020
Research Scientist / Professor
Rutgers University, Newark, New Jersey Jan 2018 - May 2020 Intelligence Analyst / Supervisor
Turkish National Police, Istanbul, Turkey Jun 1997 - Aug 2017 Data Scientist
• Developed and applied demand prediction methods like questionnaires, surveys to decrease stock level by up to 15% by using supervised learning algorithms such as Logistic Regression, Random Forest, SVM.
• Utilized price optimization models and recommended prices, which aims to increase the net profit amount by 15%
• Maintained and developed complex SQL queries and functions that meet user requirements. The Fix Solutions, Hackensack, New Jersey
Oct 2020 - Jan 2021