Reza Moshksar, PhD
Dec 2017- Data Scientist
Present International Payment Solutions, Toronto, Canada
Applied sentiment analysis and text summarization on social media data.
Predicted credit rating of clients, based on transaction records and press releases.
Settled and ranked potential customers base on geolocation and social media search history, which increased revenue by 47%.
Software & Tools
I have over 5 years of data science and professional experience in advanced analytical, statistical modeling, and machine learning. I have knowledge of Big Data plus hands-on data analysis and software development experience with frameworks like Spark and scripting languages like Python. I enjoy working collaboratively in team environments and able to work independently to complete ad hoc requests and large projects. I am one of the top 10 most liked Wikipedia users in 2017 and 2018.
Aug 2017 Rismun, Tehran, Iran
Constructed a predictive model based on the weather data. This model was part of a national project for optimizing energy consumption.
Applied to data extraction, feature engineering, predictive classification, regression modeling, model assessment, and model tuning.
Built natural language processing models to understand and extract relevant information from social media feeds and press releases. Skills
Machine Learning Deep Learning Big Data
Fraud Analytics Marketing Analytics Digital
Analytics Customer Segmentation Collection
Analytics Natural Language Processing (NLP)
Expertise in various statistical techniques:
Logistic Regression Decision Tree SVMs
Random Forest Gradient Boosting RNN
Neural Networks KNN Data mining
Python and data analytics frameworks such as:
Pandas Numpy Sklearn Tensoflow Keras
Pytorch Matplotlib Seaborn Scipy
PySpark MySQL Tableau
Ms-Office Excel RapidMiner
Sep 2013- Data Analyst
Aug 2014 Novin Dadeh Gostar, Tehran, Iran
Collected and cleaned various data types from different sources and web.
Create visualizations, reports, and dashboards
Sep 2009- Admin, Sysop, Bureaucrat, Tool and Bot developer Present
Designed and developed more than 60 live web-based bots to check user's edits, get page stats, detect anomaly edits, and list NSFW images name.
Developed bots to edit, create, and categorize pages using natural language processing (NLP).
Applied clustering technics to add category label to articles.
Identified keywords and topics of articles with 98% accuracy.
Deployed live bot and services on the cloud servers of Wikimedia. Education
Building, Environment, Science and Technology Dept. Politecnico di Milano, Milan, Italy
Building Technology Engineering, PhD
Thesis: Calibration of Building Energy Simulation through Meta-Models Publications
Calibration and uncertainty analysis for
computer models, Mar 2013, Applied Energy
Advanced Statistical Machine Learning &
Pattern Recognition Algorithms & Data
Structures Computer Vision Data Science
Specialization Deep Learning Machine
Learning Artificial Intelligence (AI)
List of Rismun company’s projects in detail
At Rismun, I was a member of the data science team where we primarily focused on data science tasks of software development.
Project, Client: New UI with a client dashboard, Mellat Bank 2017 Goal: Creating a client dashboard and developing an OMNI channel platform to allow for similar experiences across different platforms
Duty: Data science leader
• Participated in the platform development
• Identified anomalous transactions
• Predicted client credit
• Clustered clients
Achievements: The developed platform increased client transactions by 180%.
Project, Client: Intelligent Information Observation, Research and development department of Power Ministry of Iran 2016-2017
Goal: Developing a database platform for researchers to explore research articles Duty: Data science leader
• Developed a platform to find related scientific articles using public APIs such as Scopus, PatentView, and Civilica
• Crawled websites and articles through RSS feeds and google scholar searches
• Developed a word recommendation for the search box
• Created opportunities to further summarize and clustering articles, extract keywords, and recommend the most related articles
Achievements: The platform has assisted research groups to stay up to date Project, Client: Monitoring energy consumption, Iran National Engineering organization 2015-2016 Goal: Developing tools for clients to monitor and reduce household energy consumption Duty: Data scientist
Phase 1: develop a model for weather data
• Collected, cleaned, and verified weather data from different data sources
• Handled missing values in multiple variables of the weather data
• Built a predictive weather model based on the cleaned weather data using SVM Phase 2: Classify clients on their consumption behavior
• Collected and cleaned data from Gas meters
• Categorized more than 8 million clients in 10 groups based on their consumption behavior Achievements: Enabled intelligent billing, which resulted in 28% decrease in household’s energy consumption
Reza Moshksar, PhD
SQL Python Keras Pytorch Sklearn
Project, Client: Financial platform, Iran National Engineering organization 2015 Goal: Developing a paperless financial management software Duty: Data scientist
• Processed OCR results such as text cleaning, grammar, and spell checking
• Developed a model for anomaly detection and integrated it in a BI Dashboard
• Clustered city neighborhoods based on the number of building permits
• Developed models to predict future inventories in neighborhoods Achievements: Enabled the management team to detect a fraud, which saved 47% of the budget Reza Moshksar, PhD
SQL Python Sklearn PySpark