ELLA EIDLIN Raleigh, NC ***** 984-***-**** eligible to work for any us employee ad15ro@r.postjobfree.com LinkedIn MACHINE LEARNING DATA SCIENTIST COMPUTER ENGINEERING DATA Scientist Database Management Machine Learning AI Analyst Statistical Analysis Regression, Classification, Clustering Data Mining Health Informatics Computer Vision MATLAB Programming AI/machine learning Deep learning framework Image processing Research Scientist Image, and Signal Processing Linux

MacOS ML model PyTorch Python R SQL JavaScript CSS5 HTML Numpy, Scipy, pandas, and Scikit Learn anomalies MySQL Data visualization Medical Imaging X-ray imaging Image enhancement. SUMMARY

● 4+ years of experience with end-to-end Machine Learning implementation, Data Acquisition, Data Validation, Predictive Modeling, and Data Visualization.

● Good understanding of both traditional statistical modeling and Machine Learning techniques and algorithms like Regression, Clustering, Random Forest, Decision Tree, Time Series, hypothesis testing, k-means, RNN, etc.

● Experience in performing data analysis on various IDEs like Jupyter Notebook and PyCharm.

● Good knowledge of the entire Data Science process life cycle, including Data Acquisition, Data Preparation, Data Manipulation, Validation, and Visualization.

● Proficient in understanding and analyzing business requirements, building predictive models, designing experiments, testing hypotheses, and interpreting statistical results into actionable insights and recommendations.

● Good experience in generating data visualizations using Python and R creating dashboards using tools like Tableau.

EXPERIENCE

2020-2023 Bar Ilan University Research Assistent Publication:

“Unsupervised Iterative U-Net with an Internal Guidance Layer for Vertebrae Contrast Enhancement in Chest X-Ray Images” submitted to the IEEE Journal of Biomedical and Health Informatics (J-BHI).

● Developed and designed a Deep Neural Network (DNN) model for unsupervised learning, showcasing a profound understanding of the requirements for processing medical images.

● Led the development of a model from scratch, managing every stage from initial design to implementation. This encompassed not only the DNN model but also the creation of preprocessing techniques and performance evaluation protocols.

● Implemented a multi-step training strategy, refining the model iteratively to ensure optimal performance. Achieved exceptional results after careful analysis and adjustment of the training process.

● Conducted thorough analysis and preprocessing of medical data.This meticulous approach was pivotal in enhancing the model's ability to uncover subtle details in X-ray images.

● Analyzing medical imaging data, and addressing specific challenges in medical image processing. Collaborated with a team of radiologists from the Wolfson Medical Center. The collaboration involved regular discussions, knowledge exchange, and iterative model adjustments to achieve optimal results in enhancing diagnostic capabilities.

2019 - 2021 Radware Junior Data Scientist

● Conducted comprehensive analysis of network traffic patterns, identifying anomalous behavior and reducing the load on cybersecurity teams by up to 20%.

● Developed and implemented machine learning models, including Random Forest, Gradient Boosting, SVM, and clustering algorithms for real-time malware detection.

● Utilized Long Short-Term Memory (LSTM) networks for analyzing traffic sequences and capturing temporal dependencies.

● Applied statistical models and algorithms such as EWMA, Holt-Winters, PACF, and ARIMA for network traffic analysis.

2017 -2018 Intel Software developer

● Implemented Java automation and features for real-time platforms and debug tools for silicon validation.

● Created STPs (Software Test Plan) for debug tools based on face-to-face customer acceptance criteria, enhancing customer satisfaction.

● Analysis and debug multiple software tools for pre and post silicon validation and hardware platform integration.

2015-2017 Insights.US Junior front-end developer

● Developted on creating user interface features, improving the overall user experience.

● Implemented successfull migrated the front-end from Angular to React, streamlining code management.

● Developed various features, components, and functionality based on React and Angular libraries. 2014-2015 Partner Communications Ltd IT specialist

● Worked with CRM/ERP systems, deployed new hardware, and evaluated software and security risks.

● Configured and managed backup and restore procedures, ensuring the security and integrity of billing processes.

● Developed Python scripts to automate billing processes, which reduces errors and optimizing time efficiency.

TECHNICAL SKILLS

● Machine Learning: Medical image preprocessing, Feature extraction, clustering, classification, ML model, Deep Convolutional Neural Network (DCNN) architectures such as VGG16, VGG19, ResNet, U-Net, TensorFlow, PyTorch, recurrent neural network,Long Short-Term Memory (LSTM), and AI/machine learning models.

● Artificial Intelligence Algorithms (ML/AI): SVM, decision tree, random forest, evolutionary algorithm, linear regression, k-nearest neighbor, k-means, Logistic Regression, Principal component analysis (PCA), etc.

● Deploying models: Jenkins.

● Big Data and Data Science: Data acquisition, time-series data processing, data carpentry and cleaning, data processing, data analysis and real-time data processing.

● Programming Languages, Tools, and Frameworks: Python, MATLAB, JavaScript, CSS5, HTML, Linux, MacOS, LaTeX.

EDUCATION

2018 - 2023 Bar Ilan University, Israel M.Sc in Computer Science. 2010-2014 Ariel University, Israel B.Sc in Computer Science and Biology. KEY PROJECTS

● Developed from scratch Convolutional Deep Neural Netwok (CDNN) for medical image enhancement include embedded guidance layer for tiny details refining. Developed loss function oriented to recognize and sharp spinal structures in X-ray scans. The results are then evaluated on Tone Mapped Image Quality Index (TMQI), Tenengrad criterion, Entropy and LPC-SI metrics.

● Developed a medical image preprocessing method that enhances image contrast for more efficient learning. The method is based on the weighted least squares filter (WLS) and image arithmetic operations.

● Build a model to detect network threats through time series analysis. The model analyze network traffic as time serias by considering each network packet as a data point collected over time. The model based on unsupervised LSTM model is trained to capture the underlying structure of normal instances in the time series data. The discrepancy between the reconstructed (calculated by MAE loss) time series and the original input serves as the anomaly score for each instance. To classify instances as anomalies, the system introduces a threshold on the anomaly scores.

LANGUAGES

English Full professional proficiency

Hebrew Native proficiency

Russian Native proficiency

Contact this candidate