Data Scientist
Resume:
EZRA ABAH
Masters trained enthusiastic data scientist with sound demonstrated knowledge in the field’s latest technologies and best practices. Passionate about combining theoretical knowledge with actionable implementation to enable new products and solutions that solve business problems and improve customer experience. He seeks a data science role in your company where he can actively contribute to creating, managing and optimising data solutions and intelligent systems while improving in personal abilities.
Page 1 of 2
CONTACT
********@*****.***
Edinburgh, Scotland
linkedin.com/in/ezraabah
github.com/ezraabah
NON-TECHNICAL SKILLS
Business Awareness
Critical thinking for problem
solving
Innovation and creativity
Attention to detail
Exceptional Communication Skills
Teamwork and leadership
Planning and Organising
Story telling
Intellectual Curiosity
Meet deadlines
EDUCATION
Jan 2020 to Oct 2021
NAPIER UNIVERSITY [EDINBURGH, EH]
MSc. Computing with Professional
Placement (Data Science Path)
Sept 2018 to Sept 2019
HERIOT-WATT [EDINBURGH, EH]
MSc. in Oil and Gas Technology
Sept 2012 to Sept 2016
KNUST, GHANA [KUMASI, GH]
BSc. in Petrochemical Engineering
CERTIFICATION
Aug 2019 to Jan 2020
IBM DATA SCIENCE
Professional Certificate
TECHNICAL SKILLS
Programming Languages: Python, SQL, R
Machine Learning: Supervised and unsupervised ML algorithms Deep Learning: Neural Networks, CNN, RNN, Transfer Learning, hyperparameter tuning
Natural Language Processing: Word Embeddings, NLTK, Genism, Sentiment Analysis, Textual extraction, Neural language models, Text classification Database: Relational databases, NoSQL
Platforms: Jupyter Notebook, AWS Sage maker, Google Colab Data Visualization and Analytics: Tableau, Seaborn, ggplot, matplotlib, advanced Excel, structured and unstructured data analysis, time series, geospatial data analysis, Hypothesis testing, pattern recognition Version Control: GitHub
Python packages used: Pandas, NumPy, Scikit-learn, Keras, TensorFlow Cloud computing: AWS (S3 bucket, EC2), Heroku, IBM Cloud RELEVANT EXPERIENCE
Data Scientist – Edinburgh Napier University
May 2021 – Sep 2021
Actively planned and developed supervised models to predict industrial system intrusion detection at an accuracy of 99.8%. This is especially important as attacks on the system can have huge financial, quality and safety implications.
Analyzed the robustness of the model in potentially adversarial environments
Wrote SQL queries to obtain data from multiple large industrial databases to spool into CSV format and provide extracts for project stakeholders
Utilized Joins and sub-Queries to simplify complex queries involving multiple tables while optimizing procedures and triggers to be used in production
Scientific Researcher – DICON Research and Development Nov 2016 – Nov 2017
Built out the data and reporting infrastructure from scratch using Tableau and SQL to provide real-time insights into the product, marketing funnels, and business KPIs
Handled data logs and analyzed them to gain accelerated operational insight
Designed and implemented A/B experiments for products which increased sales by 7% based on findings
Collaborated with multi-disciplinary development team to identify performance improvement opportunities and integrate models Page 2 of 2
RELEVANT CERTIFICATES
COURSE COMPLETED
Python for Data Science and AI
Databases and SQL for Data
Science
Data Analysis with Python
Data Visualization with Python
Machine Learning with Python
Basic Statistics with R
HOBBIES AND INTERESTS
Playing Basketball
Playing the Piano
Self-improvement and learning
AWARDS
Winner of the University Medal,
ENU 2021
ENSA Excellence Award: Most
Inspiring Student of the year
2021
Best Program Representative,
2021 (Nominee)
PERSONAL PROJECTS
Abusive Text Classification
As covertly abusive text may be difficult to detect using traditional models, this neural language model was trained to detect both overt and covert textual abuse in tweets
Technologies used for textual cleaning and preprocessing of textual data include libraries such as Text blob, NLTK, Genism and GloVe word embedding
Model was built using an optimized single layer convolutional neural network and implemented using Keras to detect abusive text at an accuracy of 70.5%.
Insurance Claim Fraud Prediction
The model is intended to be used as a reference tool for insurance companies to help make decisions on insurance claim as traditional methods rely very heavily on manual intervention
Handled the high class imbalance in data by using the smote oversampling algorithm
Built the classification models using XGBoost, Naïve Bayes and Random Forest algorithms to obtain an optimal accuracy of 83%
Created Rest API using flask and deployed it on AWS to enable other users to make requests to the URL and receive fraud predictions as response
Employee Turn Over Prediction
Performed an extensive exploratory data analysis on HR data to understand what factors contribute most to employee turnover.
Used K-Means Clustering to identify meaningful patterns in data as well as employee traits
trained highly optimized models that predicts the likelihood that an employee will leave the company or not by up to 99% accuracy score
Suggested employee retention strategies tailored to findings Used car valuation model
This application was designed to give an estimated price for used cars considering factors such as the vehicle’s age, mileage, spec, fuel type and transmission type
The model is 100% data-driven, powered by data from thousands of vehicles and built using highly optimized random forest algorithm
Application was deployed using Heroku platform and is available for use online
EXTRACURRICULAR ACTIVITIES
Member – Big Data and Analytics Edinburgh
Aug 2021 – Date
Attend weekly seminars to discuss trending topics in analytics, data governance, cloud computing, Business intelligence, Spark, Hadoop, NoSQL etc.
Student Program Representative (MSc Computing, ENU) Jan 2020 – Sep 2020
Actively worked with staff to improve the student learning experience by negotiating solutions to problems where possible through Student-Staff Liaison Committees.
Organized and provided free tutoring to other students in data science, python programming, SQL, and R.
Contact this candidate