About Me

Data Science

&

Statistics

Hello and welcome to my website! My name is Edward Moradian. I specialize in data science and statistics. Please navigate through my website. Make sure to connect with me if you haven't already through the contact section.

Title:
Data Scientist


Location:
Los Angeles, California


Degree:
M.S. Statistcs



Resume

Tech Stack

Front-End to Back-End Development from Systems

Languages

Machine Learning Platforms
Data Engineering

Relational Databases


Data Warehouses


Dashboarding and Data Visualization
Cloud Environments


Experience

Forecasting

Numerical Forecasts on Future Dates

  • Monthly Active Users: Forecasted the number of monthly active users on a cloud gaming platform in 1-6 month forecasts.
  • Supply Chain Forecasting: Forecasted the demand amount of units based on the SKU of a refigerant product taking into account sparse, intermittent, and regularly-occurring data using different time series forecasting models.
  • Growth Curve Forecasting: Forecasted using nonlinear regression modeling fitting data to a Shifted-Gompertz distribution using an optimization method.

Causal Inference, A/B Testing

Compare Samples of Grouped Data on KPIs

  • Incremental Lift: Analyzed the incremental impact of categorical levels of store visibility on sales controlling for covariates using a multivariate regression model.
  • Causal Inference: Analyzed the incremental impact of store relocations on sales for each relocated store using a regression discontinuity design.
  • Hypothesis Testing: Tested on randomization schemes in clinical trials to determine a statistically significant difference in proportions of demographic variables of participants to check if randomization was processed correctly.

Model Interpretation

Interpreting the Results of Models

  • Model Agnostic Methods: Used PDP, ALE and ICE functions to display the percent increase in the likelihood of a positive sentiment with better or worse performing predictor variable values to determine thresholds.
  • Model Agnostic Methods: Used Shapley values at a local-level to find the influence of store visibility and other variables on store sales to discover outliers in performance.

Data Engineering, Analytics Engineering

ETL/ELT Pipelines and SQL Scripts

  • ETL Pipelines: Used airflow to develop data pipelines joining petabyte-sized core tables into a functional format for a marketing small business unit context.
  • ELT Pipelines: Developed a dbt model on a scheduled cadence to create a foundational one-big-table financial transactions model combining raw data tables from orders, fulfillments, exchanges, returns, products, discounts, and etc. for both online and retail store data from Shopify.
  • Data Orchestration: Orchestrated data pipelines to automate data ingestion and transformation through cloud functions, pub/sub topics, scheduled scripts and on-demand scripts.


Data Analysis

Descriptive Statistics on KPIs

  • Exploratory Data Analysis: Perform exploratory data analysis of model features and response variables to inform model selection, parameter and hyperparameter tuning.
  • KPI Creation: Feature engineered foundational KPIs based on absolute financial KPIs (sales, tickets), relative financial KPIs (absolute KPIs made relative to normalize for larger or smaller stores/trade areas), and interest financial KPIs (interest in the product compared to all other products).
  • Stakeholder Business Questions: Provided statistical summaries on key business questions from stakeholders and senior management.
  • KPI Creation: Modeled KPIs including lift, seasonality, benchmarks, and date period comparisons (e.g., MoM, YoY).


Data Dashboarding and Visualization

Graphically Displaying Summarized Data

  • Tableau: Developed dashboards on social media marketing program performance based on customer journey, customer benchmarks, call logs and people analytics from CRM data.
  • Power BI: Replaced manual reports with dynamic, interactive dashboards for weekly and monthly marketing reports with automated data refresh schedules.
  • Graphically represented GIS raster and shapefiles with data visualizations.




Philosophy

Favorite Philosophy Writings


John Stuart Mill

"It is better to be a human being dissatisfied than a pig satisfied; better to be Socrates dissatisfied than a fool satisfied. And if the fool, or the pig, are a different opinion, it is because they only know their own side of the question."

Karl Marx

"The philosophers have only interpreted the world, in various ways, the point, however, is to change it."

Carl Jung

"The horror which we feel for Freudian interpretation is entirely due to our barbaric or childish naivete, which believes that there can be heights without corresponding depths, and which blinds us to the really "final" truth that, when carried to extremes, opposites meet."

Heraclitus

"One cannot step into the same river twice."

Arthur Schopenhauer

"Man can do what he wants, but he cannot will what he wants."

David Hume

"Eloquence, at its highest pitch, addresses itself entirely to the fancy or the affections, captivates the willing hearers and subdues their understanding."

Plato

"The unexamined life is not worth living."

William Shakespeare

"All's well that ends well."

William of Ockham

"Entities should not be multiplied without necessity."



Contact

Edward Moradian self-portrait

Connect with me! Lets start a conversation!

Add, follow or connect with me on these sites: